In this study we compare survivorship and patient reported outcome measures in robotically assisted versus conventional Total Hip Arthroplasty (THA).

This paper investigates the hypothesis that implant survival and PROMS following THAs performed with robotic assistance were not different to outcomes following conventional THAs.

Data included all patients undergoing THA for osteoarthritis between 19 April 2016 and 31 December 2020. Analysis of PROMS outcomes was restricted to those who had completed PROMS data preoperatively and at 6 months postoperatively.

There were 157,647 procedures, including 3567 robotically assisted procedures, available for comparison of revision rates. 4557 procedures, including 130 robotically assisted procedures, had PROMS data available. The revision rate of primary THA performed with robotic assistance was not statistically different from THA performed by conventional methods (4 year cumulative percent revision 3.1% v 2.7%; HR = 1.05, p=0.67). The Oxford Hip Score, VAS for pain and the EQ-VAS score for overall health showed no statistically significant difference between the groups. The EQ-5D Utility Score showed an improved score (median score 1 v 0.88; OR = 1.58, p=0.007) for the robotically assisted group compared to the conventional group.

Robotic assisted THA was not associated with significant improvement in early revision or joint-specific PROMs. The findings may have been biased, in either direction, by unmeasured patient, surgeon, hospital and prosthesis factors. The findings (including the difference in health-related quality of life) may have also been influenced by lack of blinding. Future research should include methods to minimise these biases.

Aims

The mid-term results of kinematic alignment (KA) for total knee arthroplasty (TKA) using image derived instrumentation (IDI) have not been reported in detail, and questions remain regarding ligamentous stability and revisions. This paper aims to address the following: 1) what is the distribution of alignment of KA TKAs using IDI; 2) is a TKA alignment category associated with increased risk of failure or poor patient outcomes; 3) does extending limb alignment lead to changes in soft-tissue laxity; and 4) what is the five-year survivorship and outcomes of KA TKA using IDI?

Aims

Pelvic tilt (PT) can significantly change the functional orientation of the acetabular component and may differ markedly between patients undergoing total hip arthroplasty (THA). Patients with stiff spines who have little change in PT are considered at high risk for instability following THA. Femoral component position also contributes to the limits of impingement-free range of motion (ROM), but has been less studied. Little is known about the impact of combined anteversion on risk of impingement with changing pelvic position.

Aim

There has been a significant reduction in unicompartmental knee arthroplasty (UKA) procedures recorded in Australia. This follows several national joint registry studies documenting high UKA revision rates when compared to total knee arthroplasty (TKA). With the recent introduction of robotically assisted UKA procedures, it is hoped that outcomes improve. This study examines the cumulative revision rate of UKA procedures implanted with a newly introduced robotic system and compares the results to one of the best performing non-robotically assisted UKA prostheses, as well as all other non-robotically assisted UKA procedures.

Aims

Femoral stems with exchangeable (modular) necks were introduced to offer surgeons an increased choice when determining the version, offset and length of the femoral neck during total hip arthroplasty (THA). It was hoped that this would improve outcomes and reduce complications, particularly dislocation. In 2010, the Australian Orthopaedic Association National Joint Replacement Registry (AOANJRR) first reported an increased rate of revision after primary THA using femoral stems with an exchangeable neck. The aim of this study was to provide a more comprehensive up-to-date analysis of primary THA using femoral stems with exchangeable and fixed necks.

Bleeding related wound complications including deep infection, superficial infection and haematoma cause significant morbidity in lower limb joint arthroplasty surgery. It has been observed anecdotally that patients requiring therapeutic anti-coagulation within the peri-operative period have higher rates of bleeding related complications and those requiring intravenous heparin particularly appear to do poorly.

The aim of this study is to investigate the relationship between post-operative bleeding and wound complications in the patient requiring therapeutic warfarin, plus or minus heparin, in total hip arthroplasty surgery.

This is a retrospective cohort study reviewing 1047 primary total hip replacements performed in a single centre over a five year period and comparing outcomes of the patients on warfarin (89) with a double-matched control group of patients not on warfarin (179). Outcomes included rates of deep infection, excessive wound ooze or haematoma, superficial infection, return to OT for washout and need for revision operation. The study group was then sub analysed comparing those on IV heparin plus oral warfarin, to those on warfarin alone.

The warfarin group had significantly higher risk of deep joint infection (9% vs 2.2% p= 0.023), haematoma/wound ooze (28% vs 4% p < 0.001) and superficial infection (13.5% vs 2.2% p < 0.001) compared to the control group. In the sub analysis of the study group, those on IV heparin had significantly higher risk of haematoma/wound ooze (44% vs 28% p= 0.023) than those on warfarin alone.

The requirement of therapeutic anti-coagulation in the peri-operative period is a tenuous balance between the complications of thrombo-embolic disease and bleeding-related morbidity. In the past, perhaps the full burden of bleeding related complications has not been appreciated, but now improved understanding will enable the both the surgeon and the patient to make more informed decisions regarding therapeutic anticoagulation in elective arthroplasty surgery.

Data from recent AOANJRR shows a higher incidence of acetabular revision for dislocation of THA in uncemented compared to cemented THA (RR 1.59). We hypothesized that a difference in accuracy of component placement may be a factor. We aimed to assess any difference in accuracy between these 2 types of THA.

Patients undergoing navigated THA were prospectively recruited. Choice of uncemented or cemented THA was based on individual surgeon’s routine practice and preference and no adjustments were made for this study. All THAs (Cemented Exeter-21 and uncemented Trident/Secur fit-20) were performed through a posterior approach.

Statistical analysis: the mean and 95% confidence intervals (or median and interquartile range (IQR) for non parametric data) for each measure in both groups. ANOVA and nonparametric Mann-Whitney U test (significance level 5%). Levene’s test for homogeneity, Comparison of frequencies with chi-squared test or Fishers Exact test. Bonferroni correction where necessary.

We demonstrated a significant difference in reproducibility between components. Four of 20 (20%) uncemented cups deviated from the target inclination by 5 degrees or more compared to none of 21 in the cemented group (p=0.048). Seven of the 20 (35%) of the uncemented cups deviated from the target version by 5 degrees or more compared to none of 21 in the cemented group (p=0.003). There was a significant difference between the groups with regard to deviation from planned leg length (p< 0.001). Deviation from target leg length of greater than 5mm was found in 36.4% of the uncemented cases as compared to 8.3% of the cemented cases although due to the small numbers this was not statistically significant (p=0.16).

Statistically significant reduced accuracy of cup placement is demonstrated with uncemented compared to cemented implants. It is harder to control implant positioning in uncemented implants than cemented implants.

Introduction/Aims: Early results of a prospective randomised control trial suggested improved position of components implanted during primary hip replacement. The aim of this study is to definitively show the benefit of computer aided navigation in hip arthroplasty with regard to acetabular component position, stem position and leg length.

Method: Eighty consecutive patients were prospectively recruited. Patients were quasi-randomised, on an alternating basis, to undergo hip arthroplasty conventionally or with imageless computer navigation. Postoperatively, a CT scan was performed of the pelvis and lower limb. Using a dynamic CT planning software package, the cup and stem position was measured and compared to the position expected by the three operating surgeons in control cases and the position given by the navigation unit in the study group. Change in leg length was measured clinically and compared with the navigation predicted leg length change. Statistical analysis was performed by a statistician.

Results: Thirty nine navigated hips (29 female, 10 male) and forty one control hips (26 female, 15 male) were recruited. In the navigated group, the mean age was 65.7 and mean BMI was 29.1. In the control group, the mean age was 64.7 and the mean BMI was 29.4 in the control group. Uncemented, securfit/trident hips were used in 18 navigated cases and 20 control cases, with all other cases being cemented Exeter stems and contemporary cups. None of these differences were significant using the Mann-Whitney test. The mean operating time was 128 minutes for navigated hips and 84 minutes for controls, the difference significant at p< 0.005 using t-test.

There was no significant correlation between clinical leg length change, measured in the operating theatre and the leg length change predicted by navigation. Accuracy of cup and stem placement was assessed by comparison of the homogeneity of variances, the Levene statistic, in the navigated and control groups. The range of cup inclination, cup version and stem version was significantly narrowed in the navigation group (p< 0.05).

Conclusion: Computer navigation improves the accuracy of component placement in hip arthroplasty with respect to cup version, cup inclination and stem version with either cemented or uncemented hips.

Computer aided joint replacement surgery is being used increasingly. It is more commonly used at present in the knee replacement surgery as compared to hip replacement arthroplasty. It is still under developmental phase. The published literature shows there is increased accuracy of the component placement of acetabular cup and femoral stem. We describe the technique for the Stryker navigation system as used in total hip arthroplasty.

The technique used by us presently is an active tracker system. This is a both way communication system of infrared waves between the trackers and the sensors. The trackers are fixed to the bones, then the registration of patient specific anatomy is done and hip arthroplasty is performed with aid of the computer navigation.

The computer navigation gives the values of the component orientation in space. It gives the implant position in the pelvis and femur models generated by the computer but fed in and created by the surgeon. It is important that the data fed to the computer in making the model of pelvis and femur is accurate. It is surgeon dependent. At the end of surgery one can also evaluate impingement and range of motion. It also shows the change in offset of the centre of rotation of the hip as well as leg lengthening. While it can aid in the technical performance it is essential that the surgeon does not go blind to his operating environment as the computer navigation is to help the surgeon, not replace.

The use of computer aided joint replacement surgery is increasing exponentially. Its use in hip arthroplasty is still under developmental phase. Although the available literature shows there is increased accuracy of the component placement but there can be a number of factors on which it relies.

We have used the Stryker navigation system to aid in total hip arthroplasty for more than four years. It is improving continuously with time. Still there are many factors which are completely surgeon dependent and which can cause lot of variations in the component placement. Most important factors are the registration of patient anatomy and fixity and immobility of the bone trackers during the procedure. A number of other simple things can produce errors. We carried out some studies to see the effect on navigation values which will be presented. While use of computer navigation can aid greatly in achieving the set goals, it is dependent on surgeon thought process and appropriate implementation of the procedure.

Computer aided joint replacement surgery has become very popular during recent years and is being done in increasing numbers all over the world. The accuracy of the system depends to a major extent, on accurate registration and immobility of the tracker attachment devices to the bone. This study was designed to assess the forces needed to displace the tracker attachment devices in the bone simulators.

Bone simulators were used to maintain the uniformity of the bone structure during the study. The fixation devices tested were 3mm diameter self drilling, self tapping threaded pin, 4mm diameter self tapping cortical threaded pin, 5mm diameter self tapping cancellous threaded pin and a triplanar fixation device ‘ortholock’ used with three 3mm pins. All the devices were tested for pull out, translational and rotational forces in unicortical and bicortical fixation modes. Also tested was the normal bang strength and forces generated by leaning on the devices.

The forces required to produce translation increased with the increasing diameter of the pins. These were 105 N, 185 N, and 225 N for the unicortical fixations and 130N, 200N, 225 N for the bicortical fixations for 3mm, 4mm and 5 mm diameter pins respectively. The forces required to pull out the pins were 1475N, 1650N, 2050N for the unicortical, 1020N, 3044N and 3042N for the bicortical fixated 3mm, 4mm and 5mm diameter pins. The ortholock translational and pull out strength was tested to 900N and 920N respectively and still it did not fail. Rotatory forces required to displace the tracker on pins was to the magnitude of 30N before failure. The ortholock device had rotational forces applied up to 135N and still did not fail. The manual leaning forces and the sudden bang forces generated were of the magnitude of 210 N and 150 N respectively.

The strength of the fixation pins increases with increasing diameter from three to five mm for the translational forces. There is no significant difference in pull out forces of four mm and five mm diameter pins though it is more than the three mm diameter pins. This is because of the failure of material at that stage rather than the fixation device. The rotatory forces required to displace the tracker are very small and much less than that can be produced by the accidental leaning or bang produced by the surgeon or assistants in single pins. Although the ortholock device was tested to 135 N in rotation without failing, one has to be very careful not to put any forces during the operation on the tracker devices to ensure the accuracy of the procedure.

Introduction: Computerised navigation has been a recent addition to the armamentarium of both total hip and knee arthroplasty, following earlier experiences in neurosurgery and trauma. The first navigation devices available for total hip arthroplasty (THA) provided the surgeon with information on acetabular component placement. The current generation of hip navigation packages provide guidance with both acetabular and femoral component placement, as well as information on restoration of leg length and offset.

Method: We used the Stryker hip navigation system.

Results: We have performed in excess of 60 total hip replacements using imageless computerised hip navigation through a number of approaches including: standard posterior, mini-posterior and double-incision approaches. Our results show an increase in accuracy of component placement using computerised navigation as compared with conventional alignment guides, which is in keeping with previous cadaver and clinical studies. The accuracy of the navigation unit appears independent of incision type or size. The limiting factor with all current forms of navigation is the requirement for rigid tracker fixation to both the pelvis and the femur. Until recently, this required insertion of threaded pins and soft-tissue dissection. Loosening of pins or inadvertent pressure on a tracker may lead to false readings and the potential for inaccurate readings. The introduction of percutaneous tracker fixation has decreased problems associated with this. The streamlining of the software program, as well as the introduction of smart tools has decreased the extra time taken to 10 minutes per case.

Conclusion: The first generation of hip navigation units have been able to provide the surgeon with additional information with regard to the normal patient anatomy and allow the surgeon to more accurately place both the femoral and acetabular components. In this introductory period we have seen massive advancements with hardware (trackers and cameras have doubled in accuracy and halved in size) and the software has become more intuitive and less intrusive in the surgical process. We now await clinical studies that confirm that the use of navigation devices in total hip arthroplasty provide clinical advantages to the patient in the form of decreased dislocation rates, more accurate leg length and offset correction and better functional outcome. Navigation is being increasingly used as an adjunct to minimally invasive surgery, ensuring more accurate component position in the face of limited visibility.

Minimally invasive surgery (MIS) for THR may accelerate rehabilitation. The objectives of this study were to determine the effect of 3 surgical approaches (standard, mini (< 10cm), 2 incision Stryker approach (MIS), on length of stay, rehabilitation rates, clinical outcome, quality of life, patient safety, complications and implant position.

This study was conducted in accordance with Good Clinical Practice. Each surgeon completed 6–8 documented cases using the MIS technique before commencing enrolment to eliminate any learning curve effect. Prior to enrolment patients were assessed for eligibility and provided signed informed consent. Patient demographics, medical histories and surgical details were collected. Post-operative rehabilitation was independently documented by a physiotherapist. Clinical evaluations (HHS) were collected pre-operatively, 10 days, 6 weeks, 3 months and 1 year. Patient outcomes (SF 12/WOMAC) were collected pre-operatively and 1 year. Radiological evaluations were completed at 6 weeks. CTs/x-rays were subject to an independent review.

A sample size of 48 patients was determined based on the primary objective – length of stay. Enrolment commenced at the end of 2002 and these results are based on the first cohort of patients; based on current recruitment rates, the authors anticipate that the majority of patients will be enrolled by presentation time.

Preliminary results show mean incision lengths (cm) of 3.5/5.8 for the 2 incision MIS compared to 8.8 and 13.5 for the mini and standard respectively. Mean duration of surgery (mins) was 79 (MIS), 62 (mini) and 42 (standard). The median time (hours:minutes) from end of surgery until the first episode of knee flexion > 45°, straight leg raise, active abduction, standing, out of/in to bed, stair climbing and walking > 20 metres was shortest for MIS compared to mini and standard surgical approaches. The maximum distance walked was greatest for the MIS group. The mean length of stay (days) was shortest for the MIS group. 2.5 compared to 4.7 (mini) and 3.7 (standard).

Mean blood loss (cc) was greatest for the MIS group, 667 compared to 525 (mini) and 467 (standard). There were no intra/post-operative complications or blood transfusions.

Results suggest accelerated rehabilitation, decreased hospital stay and increased surgery duration for the MIS group. There are no safety concerns, however the procedure is felt to be quite technically demanding requiring an appropriate level of training/experience. The authors believe this is the only controlled study of this nature currently being conducted internationally.

Introduction Over recent years the techniques of femoral and acetabular impaction allografting with fresh frozen morsellised bone have become incressingly popular for revision total hip arthoplasty with osseous defects. In many centres lack of availability or legislation has required surgeons to explore alternatives to fresh frozen bone that may have different structural and biological properties. In this study we compare in vitro the load carrying capacity of irradiated morsellised bone against a control non-irradiated sample.

Methods Fresh frozen heads were divided in halves with one half irradiated at 25 kGy and the control half left non-irradiated. A custom-built pneumatic loading apparatus applied a force of 1200N at a cycle rate of 1Hz for a total of 1500 cylcles. This loading cycle was chosen to simulate the loads normally experienced by the human femur during walking gait. The reduction in height (subsidence) of each test specimen was measured and statistical analysis performed.

Results Results from each treatment group displayed similar patterns of subsidence, with an initial rapid rate of subsidence occurring up to 50 to 100 load cycles, followed by a more gradual, slower rate as the tests progressed. The results for each treatment (mean ± standard deviation) were −3.59 ± 0.91 mm and −2.98 ± 0.812 mm for the irradiated and non-irradiated groups, respectively (P+0.049). The irradiated specimens demonstrated an increased amount of subsidence compared to the non-irradiated specimens.

Conclusions This study has shown that gamma irradiation of morsellised bone allograft material decreases its load-carrying capacity, as expressed by an increase in subsidence due to an applied cyclic load. The ability for morsellised bone allograft material to bear applied loads in vivo is an important biomechanical parameter and one indicator of a successful clinical outcome. The clinical implications of this result are important when considering the most appropriate methods of treating human bone allograft material.

In relation to the conduct of this study, one or more of the authors is in receipt of a research grant from a commercial source.

Introduction Malpositioning of cup and stem in total hip replacement can result in significant clinical problems such as dislocation, impingement, limited range of motion and increased polyethylene wear. The use of mechanical alignment guides for correct cup positioning has been shown to result in large variations of cup inclination and version.

Methods Bilateral total hip replacements were performed in twelve human cadavers. While in each cadaver the operation on one side was performed with the aid of a non image based hip navigation system, the cup positioning at the contralateral hip was controlled by use of a conventional mechanical alignment guide. Post-operative cup position relative to the pelvic reference plane was assessed in both groups by the use of a 3D digitizing arm.

Results By aiming for 45° inclination and 20° anteversion for cup position the median inclination was assessed as 45.5° for the navigated group and 41.8° for the control group. Median anteversion in the navigated group was calculated as 21.9° and 24.6° for the control group. The 90 percentile showed a much wider range for the control group (36.1° to 51.8° inclination, 15° to 33.5° anteversion) than for the navigated group (43.9° to 48.2° inclination, 18.3 ° to 25.4° anteversion).

Conclusions The cadaver study demonstrates that computer assisted cup positioning using a non-image based hip navigation system allowed a more precise placement of the acetabular component in the surgeon’s desired orientation with less variance than in the control group.

In relation to the conduct of this study, one or more the authors have received, or are likely to receive direct material benefits.

Introduction Correct component positioning is critical for the stability of the prosthesis in total hip arthroplasty (THA). Malpositioning of either the femoral or acetabular component may lead to impingement or dislocation. This study aims to assess the accuracy of placement of the acetabular component in THA.

Methods Forty-six total hip arthroplasties were studied. The surgeon’s estimates of intra-operative inclination and anteversion of the acetabular component were recorded. Post-operative inclination of the acetabular component was measured from routine plain antero-posterior (AP) radiographs of the pelvis. Planar anteversion of the acetabular component was determined from AP radiographs with the beam centred over the hip using the method described by Pradhan. Planar anteversion was then corrected to ‘true’ anteversion correcting for inclination. The surgeons estimate of intra-operative cup inclination and anteversion and the radiographic position were compared.

Results The mean difference between the estimated and true cup inclination was 1.5° more than planned (range of −13° to 16°, SD 6.75). The mean difference between estimated and true values of anteversion was 1.8° less than planned (range of −32° to 25°, SD 11.07). Allowing plus or minus five degrees of error, 37% of the cups were outside the estimated inclination and 35% outside the estimated anteversion. There was a poor correlation for both inclination (Spearman’s correlation coefficient equals 0.20) and anteversion (Spearman’s correlation coefficient equals 0.25) between perceived and true cup positioning. Using the ‘safe zones’ for inclination and anteversion described by Lewinnek et al for minimising dislocation, 48% of the cups were unsafely positioned in either inclination, anteversion or both.

Conclusions Our study showed that positioning of the ace-tabular component cannot be reliably performed even by experienced surgeons. Acetabular component placement is of high importance for preventing dislocation or impingement. A tool, such as a navigation device, may be beneficial in improving acetabular component placement.

In relation to the conduct of this study, one or more of the authors is in receipt of a research grant from a non-commercial source.

Introduction Component malalignment may result in failure in total knee arthroplasty (TKA). Knee navigation systems assist surgeons with intra-operative component positioning in TKA. We report on the effect of one system on the post-operative mechanical axis of the limb and coronal alignment of femoral and tibial components in TKA.

Methods In a prospective study of 47 total knee replacements we compared 24 cases using conventional techniques to 23 cases using the Stryker Knee Navigation System. Patient groups were matched for sex, weight and age. Postoperative antero-posterior radiographs of the whole leg were used to determine the mechanical axis of the limb and coronal position of the femoral and tibial components.

Results The mean post-operative mechanical axis of the limb in the navigated group was 1.3° varus (range 7° varus to 3.5° valgus, SD=2.6). In the control group the mean mechanical axis was 0.8° varus (range 9.5° varus to 10° valgus, SD=4.4). There was no significant difference in the mean mechanical axis between the groups (p=0.6). There was no significant difference in mean coronal alignment of the femoral (p=0.99) or tibial components, (p=0.98). The 95% confidence interval for the mechanical axis was narrower for the navigated group (2.4° varus to 0.2° varus) than for the control group (2.6° varus to 1.1° valgus). Using Levene’s test (not dependant on normal distribution) the variances for the mechanical axis of the limb, and the coronal alignment of the femoral and tibial components are all significantly less in the navigated than non-navigated groups (p=0.05, 0.001 and 0.004 respectively).

Conclusions This study showed no difference in the overall mean alignment of navigated versus non-navigated knees. However, a significant decrease in the variance of alignment seen with navigation means we are seeing fewer outlying results that may lead to a decrease in mechanical failure in TKA.

In relation to the conduct of this study, one or more of the authors is in receipt of a research grant from a non-commercial source.

Introduction Malposition of the acetabular component in total hip arthroplasty is associated with a number of significant complications. In this study we acertain the accuracy of an imageless computer based surgical navigation system for positioning the acetabular component, using a mechanical hip device.

Methods A mechanical hip device was constructed that allowed accurate measurement of inclination and version, leg length and lateral offset of the acetabular and femoral components. In 31 cases these parameters were varied while a blinded operator would place the cup in the predetermined position in the mechanical hip jig. The values given by the navigation system on the screen were recorded and compared to the measurements obtained directly from the mechanical hip device.

Results The mean difference between value set by an independent person and value read by the navigation system was; inclination of the cup 0.97 ± 0 and the ante-version of the cup 1.58 ± 0.83. The leg length change contributed by the cup showed a mean difference between beforehand set value and read value on the screen of 1.48 ± 1.36 mm and for lateral offset change by the cup of 1.58 ± 0 mm.

Conclusions The accuracy of the hip navigation unit demonstrated in this study is well within the limits that would be required for clinical usage as a surgical aid in total hip arthroplasty and could provide the surgeon with a tool that enables significant improvement in accuracy for acetabular positioning.

Introduction Minimally invasive surgery (MIS) for THR may accelerate rehabilitation. The objectives of this study were to determine the effect of three surgical approaches (standard, mini [< 10 cm], Stryker two incision approach [MIS]) on length of stay, rehabilitation rates, clinical outcome, quality of life, patient safety, complications and implant position.

Methods Each surgeon completed six to eight documented cases using the MIS technique before commencing enrolment to eliminate any learning curve effect. Prior to enrolment patients were assessed for eligibility and provided signed informed consent. Patient demographics, medical histories and surgical details were collected. Post-operative rehabilitation was independently documented by a physiotherapist. Clinical evaluations (HHS) were collected pre-operatively, 10 days, six weeks, three months and one year. Patient outcomes (SF12/WOMAC) were collected pre-operatively and at one year. Radiological evaluations were completed at six weeks. CTs/x rays were subject to an independent review.

Results A sample size of 48 patients was determined based on the primary objective - length of stay. Enrolment commenced at the end of 2002 and these results are based on the first cohort of patients; based on current recruitment rates, the authors anticipate that the majority of patients will be enrolled by presentation time. Preliminary results show mean incision lengths (cm) of 3.5/5.8 for the two incision MIS compared to 8.8 and 13.5 for the mini and standard respectively. Mean duration of surgery (mins) was 79 (MIS), 62 (mini) and 42 (standard). The median time (hours:minutes) from end of surgery until the first episode of knee flexion > 45°, straight leg raise, active abduction, standing, out of/in to bed, stair climbing and walking > 20 metres was shortest for MIS compared to mini and standard surgical approaches. The maximum distance walked was greatest for the MIS group. The mean length of stay (days) was shortest for the MIS group, 2.5 compared to 4.7 (mini) and 3.7 (standard). Mean blood loss (cc) was greatest for the MIS group, 667 compared to 525 (mini) and 467 (standard). There were no intra/post-operative complications or blood transfusions.

Conclusion Results suggest accelerated rehabilitation, decreased hospital stay and increased surgery duration for the MIS group. There are no safety concerns, however the procedure is felt to be quite technically demanding requiring an appropriate level of training/experience. The authors believe this is the only controlled study of this nature currently being conducted internationally.

In relation to the conduct of this study, one or more of the authors has received, or is likely to receive direct material benefits.

Introduction Using a new surgical technique for the first time involves a ‘learning curve’. The aim of this study is to assess the technical errors encountered in the early utilisation of a computer assistance system during total knee arthroplasty and to see if this error rate decreases with experience.

Methods Thirty-two total knee replacement procedures performed by a single surgeon using the Stryker Knee Navigation system for the first time, were monitored. All technical difficulties were documented. We compared the complications encountered in the first five cases against the following 27 cases.

Results Technical difficulties related to the navigation equipment were noted in four of the 32 cases (12.5%). These included (one case each); errors in tibial pin placement, concern over initial navigation readings, pin loosening, and dropping the navigation shim plate. There was a significant decrease in technical difficulties encountered in the later cases (7%) compared to the first five cases (40 percent, p= 0.041). Additionally, in three of the 32 cases (9%) the surgeon used their clinical judgement to override the navigation readings and recut the bone, to take ligament balancing into account.

Conclusions A steep learning curve is involved when first utilising the Stryker Knee Navigation system. It is recommended that adequate training is undertaken prior to utilising knee navigation. The computer readings should be considered carefully and clinical judgement should not be overridden.

In relation to the conduct of this study, one or more of the authors is in receipt of a research grant from a non-commercial source.