Total knee arthroplasty (TKA) is an effective procedure for the management of late stage knee osteoarthritis (OA), and one which is increasing in demand. There are numerous surgical approaches of which the subvastus approach (SVa) and medial parapatellar approaches (MPa) are two options. Despite claims that SVa provides superior outcomes to the more common MPa, there has been little investigation into the physical and functional outcomes associated with each approach or their relative effect on patellar vascularity. Studies that have compared the approaches contain methodological flaws affecting confidence in the conclusions drawn. Indications for when a patient requires referral to an orthopaedic surgeon for TKA are also unclear. A better understanding of how surgical approach affects outcomes, and when a patient should be referred to an orthopaedic surgeon for review of their knee OA may optimise outcomes and improve efficiencies in patient selection.
The aims of this thesis were to address deficits in the literature concerning the relative merits of the MPa and SVa in TKA, and the clinical indications for TKA. The first aim was to conduct a systematic review of the literature to compare the two approaches to determine whether one had superior outcomes over the other. The second aim was to conduct a randomised controlled trial comparing the physical and functional outcomes associated with MPa and SVa in TKA in the short, and medium (18 month) terms. The third aim was to conduct a sub-study to determine whether the SVa maintained better patellar vascularity than the MPa, thereby reducing the risk of avascular necrosis (AVN) and anterior knee pain. The final aim was to conduct modelling to determine if physical measures, patient-assessed scoring instruments and/or clinical rating systems could be used as indicators for the timely referral of patients to an orthopaedic surgeon.
Results of the systematic review revealed insufficient or equivocal evidence supporting the SVa over the MPa. The methodological quality of most studies was poor as they either failed to randomise appropriately, adequately conceal allocation, report complications, define inclusion and exclusion criteria, or define outcomes. The use of heterogeneous outcomes prevented pooling of data for meta-analysis which may have resulted in stronger conclusions.
The randomised controlled trial was designed and conducted to address the limitations that were highlighted by the systematic review. The American Knee Society Score (AKSS) was used as the primary outcome from pre-operatively to 18 months. Secondary outcomes were knee pain, the Oxford Knee Score, three metre Timed Up and Go test, knee flexion, extension, quadriceps lag on straight leg raise (SLR), days to SLR, knee girth, length of hospital stay, operation duration, tourniquet time and surgeon perceived level of difficulty with the approach. Results of the randomised controlled trial, using linear mixed modelling for the analysis of continuous variables, revealed no difference on any outcome at any time-point between the SVa and MPa groups. The exceptions were earlier SLR in the SVa group, better AKSS Objective scores on day 1 post-operatively in the SVa group but overall better AKSS Functional scores at 12 and 18 months for the MPa group. While earlier SLR was observed in the SVa group, the surgeons perceived this approach as more difficult.
Patellar vascularity was assessed using two novel methods developed for this trial from nuclear medicine imaging techniques. These were the pat:fem ratio, which is a ratio of photon counts in the patella compared to a standardised region of interest on the femur; and the five-point Bone Vascularity Scale (BVS), which is a new quantitative method for analysing images of vascularity. There was no difference in patellar vascularity between groups on either the pat:fem or the BVS measures.
The model developed to assist primary health care providers to decide when to refer a patient with knee OA to an orthopaedic surgeon, incorporated the AKSS Functional score and knee flexion range of motion (AKSSFun/flexion) and predicted group allocation accurately 95% of the time (Odds Ratios: 1.28 to 3.40 for 1º and 5º parameter changes in flexion respectively).
This research program found, on balance, no substantive evidence supporting the superiority of the SVa over the MPa over a range of physical, functional and vascularity measures. The findings therefore refute previous supposition and claims of advantages the SVa provides over the MPa.