The use of intra-operative technological assistance has increased dramatically over the past decade, with estimates showing a 10-fold increase in utilization rates between 2008 to 2015 in the United States alone [9]. The use of robotic assistance for THA has been shown to benefit the accuracy of cup implantation, which is an important factor in ensuring favorable outcomes in the post-operative setting. The results of our investigation demonstrated that the use of a fluoroscopy-based robotic assistance system for primary THA resulted in similar accuracy and precision outcomes relative to a CT-based robotic workflow. To the authors’ knowledge, this is the first investigation which has directly compared the intra-operative performance of this specific FL-RTHA platform to that of an existing robotic system.
The first significant finding of our investigation was that both cohorts demonstrated similar average cup implantation angles and distributions for inclination and anteversion. Though this is the first investigation which has directly compared the accuracy profiles of these two robotic platforms, our findings are in line with existing literature which has demonstrated excellent intra-operative performance for both systems [10,21,25–29]. Prior studies involving the FL-RTHA platform have shown improved cup implantation accuracy relative to mTHA in a number of clinical scenarios [10,21,25,26], with comparable findings having also been reported in the CT-RTHA literature [27–29]. Furthermore, the averages and variances for cup inclination and anteversion angles obtained in this study are similar to those reported by these previous authors [10,21,25–29]. Though not statistically significant, it appeared that the FL-RTHA system trended towards greater precision for cup anteversion. This finding is especially significant, considering orthopaedic surgeons have a tendency to misjudge and subsequently misalign cup anteversion angles during impaction as opposed to inclination angles [25]. There are potentially a number of contributing factors to the improved anteversion precision in the FL-RTHA cohort, though a primary driver is likely the differences in acetabular registration workflows between systems. As discussed in Kamath et al, the FL-RTHA platform utilizes intra-operative fluoroscopic images obtained from a standard C-arm to execute pin-less digital navigation and fluoroscopy-based robotic registration of the acetabulum [21,22]. The CT-RTHA platform requires the insertion of individual navigation tracker pins, in conjunction with physical probe contact with a number of bony landmarks around the acetabular rim to perform acetabular registration [23,24]. Requiring direct probe contact with the acetabulum can prove to be challenging especially in minimally invasive DAA THA, where access to the joint space is limited. This difficulty can be further compounded by patient anatomical factors such as bony deformity and an obese body habitus [30,31]. Because image-based registration is a consistent approach that is relatively indifferent to patient anatomy [17], the FL-RTHA system may offer alternate advantages to achieve similarly precise cup implantation.
The second significant finding of this study was that the FL-RTHA and CT-RTHA platforms had similar accuracy outcomes, and also had similar proportions of cups placed within the Lewinnek zone. The authors surmise that the observed 1.2° difference in inclination accuracy between systems is a result of the FL-RTHA cohort including cases from the surgeon’s learning period with the system, which is traditionally associated with inferior surgical performance compared to the post-proficiency stage [32,33]. One investigation reported an average cup implantation angle difference of 2.8° for inclination, and 3.5° for anteversion between pre- and post-proficiency procedures when adopting an intra-operative assistance system for THA [34]. The CT-RTHA cohort only included post-proficiency procedures. Overall, it is unlikely that this small difference (1.2°) in accuracy would have meaningful clinical impact on post-operative results with regards to complication rates, functionality, or patient-reported outcomes, especially given the equal rates at which cups were placed within the Lewinnek zone in either cohort [6]. The 81.6% of cups placed within the Lewinnek zone in the FL-RTHA cohort is comparable to the 87% reported in a recent study by Stewart et al, who utilized the CT-RTHA system for DAA THA [35]. When considering the overall net benefits to the patient, this ~1° difference in accuracy should be weighed against the potential for improved cost-effectiveness [12,15], reduced operative times and pin-less workflow, and radiation exposure in the FL-RTHA cohort relative to the CT-RTHA cohort [16,17].
Our findings have several limitations. First, the retrospective nature of this study introduced the possibility of documentation biases. To mitigate this likelihood, all data were directly obtained from a rigorous review of standardized institutional electronic health records. Second, the two surgeons involved in this investigation are high-volume robotic arthroplasty surgeons, potentially limiting the generalizability of these findings to surgeons who perform lower case volumes or who have less expertise in robotic THA. Third, all FL-RTHA procedures were performed by one surgeon, and all CT-RTHA procedures were performed by a different surgeon. While both surgeons are of similar levels of arthroplasty experience, surgeon-related factors could have biased the results. To create comparable cohorts, only primary cases with a pre-operative diagnosis of osteoarthritis, avascular necrosis or rheumatoid arthritis were considered. Fourth, this was an unmatched investigation, and there were various demographic differences between cohorts that could have biased the results. In order to mitigate this limitation, we used consecutive series of patients in either cohort. However, a recent study by Barrack et al found that patient BMI and surgeon annual case volume were the two strongest predictors of cup positioning, and that patient age, sex and pre-operative diagnosis were not significant predictors of placement outcomes [36]. Lastly, cup angle comparisons could have confounded by differences in target inclination and anteversion angles between cohorts. While this may have theoretically influenced comparisons of cup placement precision, by-case adjustments of pre-operative target implantation angles for accuracy calculations would have controlled for this discrepancy. Overall, the authors do not believe that these limitations significantly biased the results.