The practice of computer-assisted orthopedic surgery (CAOS), also called navigation, has been increasing steadily for the past two decades. Numerous studies and meta-analyses have shown trends toward improved mechanical alignment in total knee arthroplasty (TKA) with CAOS.1 The outlier rate for individual bone cut measurements, as determined by deviation from the target cut, also has been shown to be reduced with navigation.2–4 Despite clear improvements in these objective metrics of success, the benefit of CAOS on clinical outcomes, as measured by functional metrics, such as the Knee Society Score, is less clear.
Adoption of navigation technology into clinical practice has been challenging. Many navigation systems are large and bulky or require proprietary tools with which surgeons may not have experience or training. Additionally, the instrumentation required for navigation can extend operating time.4 Given the need for additional training and equipment and the extended operative time, without clear clinical benefit, surgeons may be hesitant to adopt these new technologies. Although future surgical navigation systems may be truly superior to standard techniques, in the interim, there may be a role for a different class of instrumentation, navigation-enhanced instrumentation. The ExactechGPS TKA Plus navigation system (Blue-Ortho) was developed to provide real-time spatial tracking and allows for real-time computer registration and measurements, without compromising surgeon autonomy or forcing alterations to surgical technique.
The TKA Plus system was compared with conventional instrumentation as a control. In addition to assessing the accuracy of bone cuts across 6 metrics, surgeon experience and type of deformity were considered as potential influencing factors. Few studies have addressed the effects of training level on surgical accuracy in TKA. Experienced orthopedic surgeons and orthopedic surgeons still in training were recruited to determine the role that experience may play, with or without navigation enhancement. Each surgeon performed a series of cuts, with and without navigation enhancement, on varus, valgus, and neutral knees.
Materials and Methods
For this study, 2 senior surgeons (J.I.H., D.F.A.), 2 fellows (W.G.L., B.M.S.), and 4 orthopedic residents (A.B., S.K.D., P.M.L., J.P.K.) were recruited. The trial surgery consisted of distal femoral and proximal tibial resections for TKA, performed on varus, valgus, and neutral knee models (MITA Knee Insert; Medical Models). The models included a neutral knee, consisting of a neutral tibia and a neutral femur; a varus knee, consisting of a varus deformed tibia and a neutral femur; and a valgus knee, consisting of a neutral tibia and a valgus deformed femur. Experienced surgeons and fellows performed the same resections on each deformity in duplicate (6 knees total), and residents performed the same resections on each of the 3 deformities (3 knees total). The same resections were then performed with CAOS enhancement (ExactechGPS TKA Plus) on an identical set of knee models. Considering the difference in cumulative experience with TKA resection, the senior surgeons were considered experienced and fellows and residents were considered trainees for analysis purposes.
The knee models were scanned and digitized (Comet L3D [Steinbichler], Verify64 & DesignX 64 [Geomagic], and Unigraphics NX version 7.5 [Siemens PLM Software]) before and after resection. Virtual landmarks were established on the intact bone surface as anatomic references. After digitization of the preand postresection surfaces, the anatomic references were mapped onto the resected bone. Then 6 outcome metrics were evaluated: tibial slope, tibial varus/valgus, tibial depth, femoral flexion/extension, femoral valgus/varus, and femoral depth.
Each outcome metric was a planned resection with a target goal. The target numbers were 3° for tibial slope, 0° for tibial varus/valgus, 10 mm for tibial depth, 0° for femoral flexion/extension, 0° for femoral varus/valgus, and 10 mm for femoral depth. Absolute deviation of greater than 2° or 2 mm from any of these planned metrics was considered an outlier (ie, failed attempt), as defined by previous studies.3,5 Comparisons were made for all experience levels and instrumentation systems.
The primary outcome was the percentage of outliers for each outcome metric, compared between experience levels and instrumentation systems. Contingency tables comparing various test conditions were constructed and evaluated for significance by Fisher's exact test with the Statistics and Machine Learning Toolbox in Matlab (MathWorks).1 Significance was defined as P<.05.
Experienced users and trainees saw improvement with the use of navigation-enhanced instrumentation, reducing the outlier rate from 35% to 4% (P<.001) and from 34% to 10% (P<.001), respectively (Figure 1).
Comparison of outlier rates between conventional (gray) and navigation-enhanced (black) instrumentation by training level. *P<.001.
Direct comparison of each metric was made for experienced users and trainees (Figure 2). The only metric that showed a significant difference by experience level with conventional instrumentation was the tibial varus/valgus measurement. Experienced surgeons had an outlier rate of 8% compared with 63% for trainees (P=.004). No significant difference was found by experience level with the use of navigation-enhanced instrumentation for any metric in the axial or coronal plane (P>.05). Navigation-enhanced instrumentation resulted in significant improvements in accuracy across all training levels. Experienced users saw significantly improved reductions in outlier rates for femoral flexion/extension from 67% to 17% (P=.036) and for varus/valgus from 58% to 0% (P=.005). Trainees had significant reductions in outlier rates for tibial varus/valgus from 63% to 4% (P<.001) and for femoral varus/valgus from 38% to 0% (P=.002).
Comparison of outlier rates of cuts across 6 measurements obtained with conventional instrumentation for experienced users (gray) and trainees (black) (A). Comparison of outlier rates of cuts across 6 measurements obtained with navigation-enhanced instrumentation for experienced users (gray) and trainees (black) (B). Absence of a measurement bar indicates an experimental outlier rate of 0%. *P<.05.
Assessment across experience levels and instrumentation types showed no significant differences in outlier rates for valgus or varus deformities compared with neutral alignment (P>.05; Figure 3). A significant reduction in total outlier rate for coronal alignment occurred with navigation-enhanced instrumentation compared with conventional instrumentation of 32% to 11% for neutral alignment (P=.004), 44% to 4% for valgus (P<.001), and 26% to 10% for varus (P=.016).
Outlier rate by deformity type, across all training levels, for each instrument. Neutral (light gray), valgus (dark gray), and varus (black). *P<.05.
The development and implementation of navigation-enhanced instrumentation systems will continue to affect orthopedic surgery, especially total joint arthroplasty. Improvements in precision and accuracy made possible by real-time spatial tracking potentially can improve the accuracy and reproducibility of surgery. Computer-assisted orthopedic surgical devices have been shown to reduce outlier rates by more than 50%, and the same meta-analysis showed that TKA complication rates were reduced from 7% with traditional instrumentation to 4% with navigation systems.1 A major challenge in the implementation of these devices has been the additional training and burden placed on surgeons to incorporate the device into their surgical technique. Some navigation systems impose physical constraints that force modifications to technique that may have no beneficial effect on outcomes. Navigation-enhanced instrumentation is a promising step that augments surgical accuracy without modifying surgical technique to focus on fiduciary registration.
Interestingly, experienced orthopedic surgeons and those still in training (ie, fellows and residents) had similar outlier rates for 5 of 6 metrics. Except for tibial varus/valgus with conventional instrumentation, increased surgical experience alone did not produce a significant difference in the rate of outliers. This finding argues that the outlier rate with repeated use of conventional instrumentation does not improve over time. This is counter to the authors' bias as surgeons and suggests that surgeons simply may become more confident with the same level of imprecision over time rather than actually improving from a nominal level of inaccuracy that is determined by hand-eye coordination, manual skills, and conventional instrumentation available in their training and practice. However, use of a navigation-enhanced instrumentation system, with no previous exposure or additional training, decreased the rate of outliers, independent of experience level.
Overall, the results provide insight into modifiable factors that account for surgical accuracy. Experience level does not correlate with improved outlier rates, nor does experience confer an advantage for most cuts. Neither experience level nor deformity had a significant effect on outlier rates. It appears that the type of instrumentation used is the only modifiable factor in significantly influencing the rate of outliers. This finding suggests that the effect of surgical training and experience on reducing outliers is limited and that improving instrumentation systems offers a chance to improve accuracy across an array of surgical experience levels and anatomic alignments.
Improvements seen with enhanced conventional instrumentation may or may not be a result of navigation alone. This particular navigation system effectively requires the surgeon to make an initial measurement with conventional instrumentation and then immediately reassess that measurement with navigation. This design capitalizes on a common sense principle: measure twice; cut once. Future investigations should consider the possibility that navigation-enhanced instrumentation may be distinct from navigation and conventional instrumentation, both of which measure only once.
There are several limitations to the conclusions that can be drawn from this study. The goal of this study was to gain a broad perspective on the effects of navigation-enhanced instrumentation. Therefore, some effects have insufficient statistical power to be considered significant. This study was likely highly specific in that the effects seen are likely true, at the expense of low sensitivity to effects of smaller magnitude. Nonetheless, these results provide insight into potentially viable future investigations. Additionally, measurements of bony cuts and outlier rates were used as surrogate markers for surgical success, but no clinical outcome measures were assessed, and as an in vitro model study, perioperative component positioning may not equate to final positioning.6
This trial was conducted broadly in an attempt to capture and specify which metrics and conditions affect outlier rates. Surgical experience level and anatomic deformity play a relatively minor role in the rate of surgical outliers. The authors suggest that the choice of instrumentation, specifically, navigation-enhanced instrumentation, plays a far more profound role than previously recognized. Navigation alone suppresses outliers but requires deviation in surgical technique and has had limited use over the past decade. Navigation-enhanced instrumentation predictably improves the outlier rate across all training levels, independent of bony deformity, during TKA. Navigation-enhanced instrumentation is a promising and novel hybrid system of instrumentation, with minimal deviation from conventional surgical technique and a “measure twice; cut once” effect that may extend beyond navigation.
- Nam D, Cody EA, Nguyen JT, Figgie MP, Mayman DJ. Extramedullary guides versus portable, accelerometer-based navigation for tibial alignment in total knee arthroplasty: a randomized, controlled trial. Winner of the 2013 HAP PAUL award. J Arthroplasty. 2014;29(2):288–294. doi:10.1016/j.arth.2013.06.006 [CrossRef] PMID:23871707
- Mason JB, Fehring TK, Estok R, Banel D, Fahrbach K. Meta-analysis of alignment outcomes in computer-assisted total knee arthroplasty surgery. J Arthroplasty. 2007;22(8):1097–1106. doi:10.1016/j.arth.2007.08.001 [CrossRef] PMID:18078876
- Jenny JY, Boeri C. Computer-assisted implantation of total knee prostheses: a case-control comparative study with classical instrumentation. Comput Aided Surg. 2001;6(4):217–220. doi:10.3109/10929080109146086 [CrossRef] PMID:11835617
- Schmitt J, Carsten H, Kienapfel H, et al. Navigation of total knee arthroplasty: rotation of components and clinical results in a prospectively randomized study. BMC Muskuloskelet Disord. 2011;12:16. doi:10.1186/1471-2474-12-16 [CrossRef] PMID:21235810
- Ferrara F, Cipriani A, Magarelli N, et al. Implant positioning in TKA: comparison between conventional and patient-specific instrumentation. Orthopedics. 2015;38(4):e271–e280. doi:10.3928/01477447-20150402-54 [CrossRef] PMID:25901619
- Shi J, Wei Y, Wang S, et al. Computer navigation and total knee arthroplasty. Orthopedics. 2014;37(1):e39–e43. doi:10.3928/01477447-20131219-15 [CrossRef] PMID:24683655