Abstract
Introduction
The advantages of the direct anterior approach (DAA) for total hip arthroplasty include the preservation of external rotators and hip abductors thus leading to quicker recovery times. To our knowledge, there is no objective method in the literature to predict the level of difficulty for femoral exposure through the DAA. It would be beneficial to the surgeon learning the DAA to assess difficulty pre-operatively to avoid prolonged operative times. The purpose of this study was to develop a predictive model of femoral exposure difficulty in the DAA using a combination of demographic data and radiographic measurements.
Methods
305 post-operative radiographs of consecutive THA's in patients (184 female, 120 male) with primary or secondary osteoarthritis, mean age 64.6 (range 26–91, SD=11.43) performed through the DAA by one of the co-investigators from 12/2005 to 12/2009 were retrospectively reviewed by two separate observers. The observers were blinded to the difficulty level of femoral exposure. Standard post-operative AP pelvis films were assessed with TraumaCad software (TraumaCad 2.2, Voyant Health, Columbia, MD) to make radiographic measurements as shown in Figure 1–2. Each radiograph was calibrated using the size of the femoral head implant. Exclusion criteria included films that had inadequate coverage of the entire pelvis, mal-rotation, or poor exposure. Statistical analysis was performed using STAT 9.1 (StatCorp; College Station, Texas, USA). A two-sided Kruskal–Wallis test was utilized for non-parametric data. Chi-squared tests and Fisher's Exact Test were used to compare proportions. Statistically significant associations were then added to a multivariate model predicting an outcome of difficult exposure.
Results
The difficult exposures were equally distributed throughout the study period. The side of the THA was not associated with a difficult exposure (χ2=0.5516, p=0.968) whereas 66% of difficult cases were male (χ2=38.5323, p=0.0001). Height, weight, BMI, and age were all independent predictors of a difficult exposure, with taller (>175cm) more difficult than shorter (p=.0001), heavier (>100kg) more difficult than lighter (p=.0001), higher BMI (>32) being more difficult than lower BMI (p=.0001), and younger age (<60) being more difficult than older age (p=.003). Radiographic criteria that were predictive of difficult femoral preparation were decreased distance (<110mm) between teardrop signs (p=.0001), increased distance (>211mm) between each SLA (p=0.013), and increased distance (>306mm) between the GT (p=.007). The distance between each ASIS (p=0.375), ASIS to GT (p=.191), and ASIS to SLA (p=.191) were not predictive of difficult femoral preparations. From this, we determined a simple pre-operative scoring tool which allows the surgeon to predict difficult femoral preparations with an 87% sensitivity and easy preparations with >95% specificity.
Conclusion
The DAA approach has proven difficult to learn for many surgeons. Careful patient selection can facilitate the learning curve and improve patient outcomes. We describe a simple to implement preoperative rating scale, which gives the surgeon learning DAA an algorithm for appropriate patient selection. Selecting the appropriate patient can reduce the risks to the patient and minimize the cost to society of integrating new surgical techniques.
