Curr Rev Musculoskelet Med (2017) 10:378–387

DOI 10.1007/s12178-017-9415-6

QUALITY AND COST CONTROL IN TJA (B WADDELL, SECTION EDITOR)

Technology in Arthroplasty: Are We Improving Value?

Bradford S. Waddell 1,2 & Kaitlin Carroll 3 & Seth Jerabek 3

Published online: 7 July 2017

# Springer Science+Business Media, LLC 2017

Abstract studies evaluating the direct cost and value of these

Purpose of Review Total joint arthroplasty is regarded as a technologies.
highly successful procedure. Patient outcomes and im- Summary Navigation and robotics have been shown to im-
plant longevity, however, are related to proper alignment prove component position in total joint arthroplasty, which
and position of the prostehses. In an attempt to reduce can improve patient outcomes and implant longevity. These
outliers and improve accuracy and precision of compo- technologies offer a promising future for total joint
nent position, navigation and robotics have been intro- arthroplasty.
duced. These technologies, however, come at a price.
The goals of this review are to evaluate these technologies Keywords Technology . Arthroplasty . Value . Navigation .
in total joint arthroplasty and determine if they add value. Robotics . Cost
Recent Findings Recent studies have demonstrated that navi-
gation and robotics in total joint arthroplasty can decrease
outliers while improving accuracy in component positioning. Introduction
While some studies have demonstrated improved patient re-
ported outcomes, not all studies have shown this to be true. Total hip and knee arthroplasty are highly successful opera-
Most studies cite increased cost of equipment and longer op- tions that offer significant pain relief and restore function in
erating room times as the major downsides of the technologies most cases [1, 2, 3]. Studies have shown that proper alignment
at present. Long-term studies are just becoming available and in both hip and knee replacement can lead to improved post-
are promising, as some studies have shown decreased revision operative outcomes including postoperative function, postop-
rates when navigation is used. Finally, there are relatively few erative pain, and improved longevity of the implant [4•, 5•, 6].
Improving outcomes after total joint replacement have al-
ways been a goal for orthopedic surgeons. In the pursuit of
improving outcomes after surgery and longevity of the im-
This article is part of the Topical Collection on Quality and Cost Control plants, surgeons have begun to use technology in hopes of
in TJA achieving these goals.
Computer-navigated and robotic-assisted total joint re-
* Bradford S. Waddell placement has been introduced over the last two decades. As
Bradford.Waddell@ochsner.org with any new technology, these systems come with a learning
curve and a potential increased cost to the surgical procedure.
1
Department of Orthopaedic Surgery, Ochsner Medical Center, 1514 Increased initial cost can be offset by reduction in hospital
Jefferson Highway, New Orleans, LA 70121, USA stay, complications, and future revision surgeries. The goals
2
Ochsner Clinical School, University of Queensland School of of this review are to provide a brief overview of these tech-
Medicine, New Orleans, LA, USA nologies while evaluating the literature surrounding these
3
Department of Orthopaedic Surgery, Hospital for Special Surgery, technologies and finally to review studies related to value of
New York, NY, USA technology in total joint arthroplasty.
Curr Rev Musculoskelet Med (2017) 10:378–387 379

Total Knee Arthroplasty

Total knee arthroplasty has enjoyed long-standing success in

pain relief and improvement of function. These two outcomes,
along with longevity of the implant, however, have been
shown to be related to correct positioning of the components
[4•, 6–8]. Navigation has been introduced in total knee
arthroplasty in an attempt to improve component position
and restore the mechanical axis.

Computer-Assisted Total Knee Replacement

Computer-assisted surgical (CAS) navigation was intro-

duced in the 1990s. CAS total knee replacement comes
in three distinct types. These include imageless (includ-
ing accelerometer based), preoperative image based, and
intraoperative image based. Depending on the types of
navigation, the computer integrates the information from
landmarks on the images and/or landmarks taken at the
beginning of the surgery with data acquired during the
surgery to determine the frontal and sagittal plane to Fig. 1 KneeAlign accelerometer-based knee navigation (OrthoAlign
properly position the cutting guides. Aliso Viejo, CA) (Courtsey of Seth Jerabek)

Imageless CAS
has been created. The image-based approach gives the
Two main types of imageless systems exist: accelerometer surgeon the ability to plan implant positioning before sur-
based and optical navigation systems. The accelerometer- gery. Further, the surgeon has the ability to intra-
based (Fig. 1) navigation uses a hand-held accelerometer operatively alter the position of the components because
that attaches to the bone to establish the axis of the limb there is no constraint to the specific plan.
and then attaches to the cutting guides to guide its posi-
tion. Optically based imageless navigation systems
(Fig. 2) use optical localization between probes and se-
cured pins that define the surface of the bone or bony
landmarks and an optical sensor attached to the computer.
Imageless systems have the benefit of avoiding radiation
with the preoperative computerized tomography (CT) and
avoid the cost of preoperative CT or magnetic resonance
imaging (MRI). The accelerometer-based system has an-
other added benefit over older imageless and image-based
navigation systems as they avoid securing pins to the pa-
tient outside the immediate surgical field. Though rare,
pin sites have been shown to be susceptible to infection
or fracture at the pin site [9]. New technologies are mov-
ing towards pinless navigation to avoid these complica-
tions [10, 11].

Image-Based CAS

Image-based navigation systems use either a pre-operative

CT or MRI scan to form a 3D model in the computer of
the patient’s specific anatomy. Then, prior to beginning
the surgery, the surgeon uses optical navigation to register Fig. 2 Imageless navigation for total knee arthroplasty (Exactech GPS
the patient which is then matched to the 3D model that Primary Knee Gainesville, FL) (Courtesy of Brad Waddell)
380 Curr Rev Musculoskelet Med (2017) 10:378–387

Outcomes average were about 5 years younger than conventional pa-

tients (p < 0.00001). Further, they found that the navigated
As previously mentioned, the ultimate goal of adding comput- group had a mechanical axis within the acceptable 3° of neu-
er assistance to total knee replacement is to add value by tral in 78% of the cases, whereas conventional was in the
improving patient-reported outcomes and reducing surgical acceptable range in 58% of cases. At 10 years, 19 revisions
costs by reducing complications and the need for revision had been performed, with 8 for aseptic loosening. There were
surgery. In their randomized control trial of patients undergo- 7 in the conventional group and 1 in the CAS group, equating
ing simultaneous bilateral total knee arthroplasties, Zhang and to 87% survival for the conventional group and 98% in the
colleagues found that the CAS total knee group had signifi- CAS group (p = 0.03). Using the HSS and KSS scores for
cantly better coronal and sagittal alignment compared with the postoperative outcomes, they found no difference between the
conventional total knee group. In the CAS group, there were two groups at 10 years (p > 0.19). The authors do comment
no outliers greater than 3° from the mechanical axis in the that those patients undergoing CAS surgery had a lower
coronal plane, whereas there were 9 in the conventional group. “pain” category than the average population. They conclude
The two groups did not differ in terms of rotation of the femur, that CAS offers a higher chance of optimal component posi-
and at early follow-up (6 months), there was no difference in tion and a significantly lower revision rate at 10 years.
outcomes based on Hospital for Special Surgery (HSS) In another mid- to long-term study, de Steiger used the
Scores. Important for the discussion of cost, the CAS group Australian registry to assess revision rate in total knee
had an average operative time of 90 versus 58 min in the arthroplasty with and without the use of CAS [16]. They eval-
conventional group. In another study which utilized an uated the registry from January 1, 2003 until December 21,
accelerometer-based navigation technique, Nam and co- 2012 and found that the rate of CAS surgery increased from
authors used postoperative radiographs to determine the accu- 2.4 to 22.8%. They found that the overall revision rate among
racy of the system in placing the tibial component. They found conventional total knees during that time period was 5.2%
this navigation technique to accurately place the tibial compo- compared with 4.6% in the navigated group (p = 0.15). This
nent within 2° of the coronal (90°) and sagittal goal (3° slope) difference became even greater when comparing knees per-
97.6 and 96.2% of the time, respectively [12]. Short-term formed in those patients under 65. In patients under 65, there
outcomes in total knee replacement have also been shown to was a statistically smaller revision rate in the CAS group, 6.3
be improved with CAS. In their meta-analysis, Rebal et al. versus 7.8% (p = 0.011). Further, when specifically evaluating
found that CAS was more likely to place the components revision for aseptic loosening in those patients less than 65,
within 3° of the ideal mechanical alignment compared with again, CAS had a significantly lower revision rate (1.6 vs
conventional total knee arthroplasty (87.1 vs 73.7%, p < 0.01). 2.6%, p = 0.001). Finally, they evaluated the rates of major
Further, they found navigated knees to have increased Knee revision and found that in the entire age cohort, CAS de-
Society Scores (KSS) at both 3-month (68.5 vs 58.1, p = 0.03) creased the rate of major revision over conventional surgery
and 12–32-month follow-up (53.1 vs 45.8, p < 0.01). Similar (2.1 vs 2.7%, p < 0.001). They found no difference in the rates
to other studies, their analysis also demonstrated that CAS had of minor revision between the entire cohort with regards to
significantly longer operative times over conventional surgery navigation.
(101.6 vs 83.3, p < 0.01). Longer term follow-up is important Another potential benefit of CAS in knee arthroplasty is
in determining if CAS is in fact adding value to knee replace- mitigating the need to violate the femoral canal, which can
ment. Cip and co-authors demonstrated an improvement in decrease blood loss. In their study, Licini and Meneghini dem-
Insall knee Score and HSS knee score at minimum 5-year onstrated that total knee arthroplasties performed with com-
follow-up. They did not, however, find a difference in puter navigation had less hourly hemovac drain output
Western Ontario and McMaster Universities Arthritis Index (p = 0.02), smaller hemoglobin change (p = 0.001), and a
(WOMAC) scores [13•]. In a separate study, at an average lower estimated blood loss (p = 0.001) [17]. They proposed
of 46 months (range 30–96 months), Blakeney and colleagues that avoiding violating the femoral canal as a potential source
found a trend towards improved Oxford Knee Scores (OKS) of less blood loss. In a meta-analysis by Moskal and col-
in the computer-navigated group and a significantly improved leagues, they reviewed studies with a total of 7151 knee
OKS when the mechanical axis was within 3° of neutral arthroplasties and found that blood loss was significantly low-
(p = 0.045) [14] (more often seen in the navigated group). er in the navigated group, along with improved component
A recent study assessed outcomes at 10-year follow-up. alignment and higher clinical ratings [18].
This study, performed by Baumbach and collegaues, retro- Not all studies demonstrate benefit with CAS in knee re-
spectively reviewed a consecutive series of 217 cases and placement surgery. In 2013, Burnett and Barrack performed a
were able to report on 46 conventional and 50 CAS total knee systematic review of navigated versus conventional total knee
replacements at 10 years [15]. They found that the CAS group replacement asking the question: “Does the literature contain
had a 22-min longer average surgery (p < 0.0001) and on evidence of better long-term function and lower revision rates
Curr Rev Musculoskelet Med (2017) 10:378–387 381

with navigated TKA compared with conventional TKA?”

[19]. They conclude that navigation does improve coronal
plane alignment, but found little evidence of improvement
with regards to other variables. They found longer surgical
times and unique complications (such as pin site fracture and
infection) associated with CAS. At the time of their study, they Fig. 3 Navio Knee System (Courtesy of Brad Waddell)
concluded that current studies did not support CAS over con-
ventional TKA. allows the surgeon to preoperatively template both the bone
cuts and the implant size and position to be used in the surgery.
Robotic-Assisted Total Knee Arthroplasty While still early in its release, it has been shown to be more
accurate than conventional TKA [24] and more friendly to soft
Another technology introduced in knee replacement surgery is tissues [25]. Future clinical studies are needed to determine
robot-assisted surgery. At present, four robotic systems are the value added by all of these robotic TKA systems.
available for knee replacement.
The Robodoc/TSolution One Surgical System (Curexo
Technology, Freemont, CA and Think Surgical Inc.) is avail-
able for total knee arthroplasty. The system uses a preopera- Unicompartmental Knee Arthroplasty
tive CT scan and a milling reamer to prepare the tibia and
femur in TKA. While not in widespread use in the USA, While not a principle topic of this article, it is necessary to
Robodoc has been shown to improve accuracy compared with mention the role navigation and robotic assistance has played
conventional total knee arthroplasty [20, 21]. in improving outcomes and component alignment with
iBlock is a robotic cutting guide from OMNIlife Science unicompartmental knee arthroplasty (UKA). UKA is on the
(East Taunton, MA). This system is a motorized bone rise and currently comprises about 8% of all knee
mounted cutting guide that uses an intraoperatively created arthroplasties and the rate of use is increasing [26].
3D model of the patient’s bone from data taken by the sur- In a prospective, randomized controlled study from the
geon. The robotic-motorized cutting block is mounted to the United Kingdom, Bell and colleagues compared 58 conven-
patient’s femur, and after the sizing and bone cuts are deter- tional UKAs with 62 UKAs performed with robotic assistance
mined from the computer model, the guide moves to allow the (MAKO Surgical at the time of study, currently Stryker Corp.
surgeon to make the specific cuts. This system is used in Mahwah, NJ). They found that accuracy of component posi-
conjunction with the Nanoblock, which is a separate adjust- tioning was significantly improved with the robotic arm assis-
able tibial cutting guide. Few studies exist using this system; tance (p < 0.01) [27]. In another study using the robotic arm,
however, the system has been shown to be extremely precise Pearle et al. followed 909 UKAs for 2.5 years, finding a revi-
and accurate in a saw bone model [22]. In the single clinical sion rate of only 1.2%, and in those not revised, they have
study available, it allowed a single surgeon to stay within 3° of 92% satisfied or very satisfied [28]. Using the Navio system
neutral with regards to bone cuts in the first 100 cases [23]. described above in a cadaveric study, Lonner et al. demon-
The Navio surgical system (Smith and Nephew Memphis, strated less than 2° of error when implanting unicondylar
TN) (Fig. 3) is somewhat of a robotic system, but is more
accurately labeled a hand-held, imageless burring system that
utilizes an intraoperatively created 3D model from the pa-
tient’s anatomy. The system then allows the surgeon to plan
the bone resection and implant sizes prior to beginning the
bone resection with the burr tool. The system tracks the pa-
tient’s limb and the hand-held burring tool, stopping or
retracting the burr to keep the surgeon within the defined
limits of the implant resection. Navio began with limitations
to only UKA but has now been expanded to total knee
applications.
The final robotic system for TKA is the Robotic Arm
Interactive System (Rio; Mako Stryker, Mahwah, NJ)
(Fig. 4). Only just recently released for total knee application,
this system uses a robotic-controlled arm with saw blade at-
tached to make the bone cuts in TKA. The patient undergoes a Fig. 4 Mako RIO Robotic-Assisted Knee Replacement (Courtesy Brad
preoperative CT scan from which a 3D model is created. This Waddell)
382 Curr Rev Musculoskelet Med (2017) 10:378–387

prostheses with the Navio Precision Freehand Sculpting tool experienced surgeons without navigation [32]. Similarly, oth-
(Blue Belt Technologies at the time of study, currently Smith er authors have shown that over 97% of acetabular compo-
and Nephew, Memphis TN) [29]. nents placed with imageless navigation were within the safe
zone of ±10° for both inclination and anteversion [33].
A primary benefit of computer navigation for THA appears
Total Hip Arthroplasty to be a reduction in the number of cups placed far outside the
acceptable safe zone. A prospective RCT comparing conven-
Imageless Navigation tional non-navigated THA to the ORTHOsoft imageless nav-
igation system (Zimmer, Warsaw, IN) demonstrated no differ-
Non-robotic computer navigation systems can be categorized ence in cup abduction angles, but final cup anteversion devi-
based on whether they require pre-operative imaging. ated significantly less from the planned angle of 15° in the
Imageless navigation relies only on intra-operative registra- navigated group [34•].
tion of bony landmarks to create a virtual 3D model of the Imageless navigation can also facilitate limb length resto-
patient’s anatomy and determine the patient’s relative posi- ration. In a randomized comparison of imageless navigation
tioning. Whereas image-based navigation systems can gener- and fluoroscopy (without navigation), there were no signifi-
ate patient-specific 3D reconstructions of the patient’s actual cant differences between groups in leg length restoration and
anatomy, imageless systems can only map landmarks identi- femoral offset, but the navigated group had fewer outliers
fied by the surgeon onto a generic pelvis model. These sys- more than 5 mm outside the target zone accounting for both
tems use specific intraoperative landmarks to allow the com- leg length and femoral offset [35]. Other groups have also
puter to provide values for version and inclination, offset, and demonstrated the ability to restore limb length to within
leg length. This can be one of the limitations of this technique, 6 mm of the contralateral limb in over 95% of cases [36],
as proper registration is key to success. Imageless navigation although there is currently no clear evidence that navigation
requires less capital investment than in robotic equipment, restores limb length better than conventional THA.
spares the patient radiation exposure and expense associated Renkawitz et al. conducted an RCT comparing convention-
with pre-operative imaging, and requires only minimal set-up al THA to a femur-first technique using the Brainlab
for each surgical case. Although surgical time decreases with imageless navigation system, which presented the surgeon
experience, imageless navigation typically lengthens the total with a 3D representation of the recommended cup position
surgical time by about 12–18 min, due to additional registra- to maximize bony coverage and impingement-free motion.
tion steps. Both groups had over 87% bony surface contact with the
OrthAlign (OrthAlign Inc., Aliso Viejo, CA) is a dispos- cup, but more patients in the navigation group achieved max-
able palm-sized accelerometer-based device that is compatible imal impingement-free range of motion (84%, 48/66 vs 65%,
with all implant systems for both hip and knee arthroplasty 43/69). Harris hip scores were significantly higher in the nav-
[30]. The device consists of a disposable computer display igated group at 6 weeks, but the difference was clinically
unit and a reference sensor, to be used for acetabular prepara- unimportant, and by 6 months and 1 year, there were no dif-
tion. Rather than pointing to multiple reference points on the ferences between groups [37]. Patient satisfaction, clinical
limb to define the femoral reference plane, the surgeon moves outcomes, and manual ROM testing were equivalent in both
the limb in specific patterns. The computer then calculates the groups at 1 year. Retrospective comparisons of imageless nav-
mechanical axis of the limb based on measurements from the igation and conventional THA found no differences in Haris
accelerometer. The device is then mounted on cutting jigs, and hip scores, periprosthetic bone mineral density, range of mo-
it provides real-time feedback for the surgeon to perform bony tion, or polyethylene wear at 5–7 years postop [38].
resection [30].
Another imageless system is the HipXpert (Surgical Image-Based Navigation
Planning Associates, Medford, MA) that enables a simple
mechanical device to dock to the pelvis in a patient-specific Image-based navigation uses pre-operative CT, MRI, or fluo-
manner and thereby guide cup orientation. Cup anteversion roscopy to facilitate surgical planning and execution. CT-
and inclination were significantly more accurate in cups guided navigation is the most common form of image-based
placed using the HipXpert system than in those placed with navigation. Pre-op planning for non-robotic CT-based systems
traditional CT-based navigation [31]. is essentially the same as for CT-based robotic systems.
Research has confirmed that imageless navigation systems Intraoperatively, the surgeon registers bony landmarks and
are generally precise and reliable. In one study, imageless instruments and receives computer feedback about instrument
navigation yielded precise and reproducible cup positioning and implant positioning. However, in contrast to robotic THA,
within 5° for both inclination and abduction, compared to 12° the surgeon executes the entire procedure without any con-
for inclination and 13° for abduction among cups placed by straint from the robot. This gives the surgeon more freedom
Curr Rev Musculoskelet Med (2017) 10:378–387 383

to alter the preoperative plan based on intraoperative findings, computer based on the 3D imaging preoperative plan. For
but also allows the surgeon to err or place the components example, a semi-active robot such as the MAKO Robotic
outside the recommended zone. Arm Interactive Orthopedic (RIO) System (Stryker,
Fluoroscopic navigation is similar to imageless navigation Mahwah, NJ) will permit the surgeon to slightly adjust the
because neither uses preoperative imaging. Instead, the sur- angle at which the cup is reamed, but will not permit the
geon registers each landmark intraoperatively using fluoros- surgeon to push the reamer more than a few degrees beyond
copy. Although fluoroscopic navigation may seem more ac- the planned cup position. By combining 3D femoral and ace-
curate because it includes patient-specific imaging, the regis- tabular planning, the surgeon can plan modifications to the
tration process is cumbersome and it does not appear to offer femoral stem size, offset and positioning to account for chang-
any advantages over imageless navigation [39]. es to the acetabular center of rotation. The software measures
CT-based navigation systems provide more accurate mea- changes in limb length and combined offset relative to both
surements of cup alignment than conventional THA, resulting the preoperative ipsilateral limb and the contralateral limb.
in fewer cup positioning outliers [40, 41, 42•]. One retrospec- Because robots do not fatigue and can easily overcome torque
tive review of 180 navigated THAs and 120 manual THAs from larger reamers, robots permit single-stage acetabular
demonstrated a significantly lower rate of cup placement out- reaming, which can decrease surgical time.
side the Lewinnek zone (0 vs 26%) and significantly fewer Surgical time is consistently increased in robotic THA,
postoperative dislocations (0 vs 8%) in the navigated group, although the reported increase in surgical time varies widely
although there was no significant difference in 13-year im- between authors [47–49]. However, with single-stage reaming
plant survival [43]. A systematic review of publications in- and familiarity, robotic hip surgery can become time neutral.
cluding 400 patients revealed no significant difference in Robotic systems require substantial upfront financial in-
mean cup inclination or anteversion between the conventional vestment for the robot and software, as well as recurring costs
and navigated groups, but variability in cup position and the associated with each surgical case. At one Japanese institu-
risk of placing the cup outside the safe zone were significantly tion, each robotic surgery incurred $1500 additional cost for
reduced in the navigation group [44]. disposable equipment such as drapes and bone cutters [47].
Robotic THA is also subject to the disadvantages of CT-based
Robotic Navigation systems, including increased cost and radiation exposure as-
sociated with the scan and longer time devoted to pre-
Robotic systems assist the surgeon in executing the surgery by operative planning.
transferring the pre-operative imaging data and templating to a
robotic surgical assistant with an articulating arm that attaches
to the surgical instrument or implant, such as the acetabular Outcomes
reamer or cup [45] (Fig. 5).
Robotic assistants can be classified into “fully active” and Studies demonstrate some improvements in femoral compo-
“semi-active” systems, depending on the degree to which they nent radiographic parameters associated with the fully active
permit the surgeon to retain some control over the task. Drs. ROBODOC system. The initial randomized multicenter fea-
William Bargar and Howard Paul created the first robotic de- sibility study for ROBODOC using a posterior approach dem-
vice for total joint arthroplasty, the ROBODOC system, in onstrated statistically significant improvements in fit, fill and
1985 [46]. It is a CT-based computer-aided robotic milling alignment compared to non-robotic THA, and no ROBODOC
device originally designed to facilitate femoral component patient sustained an intra-operative fractur [50, 51]. A
preparation and implantation. Fully active robotic systems like Japanese RCT comparing ROBODOC and conventional
ROBODOC can perform specific tasks or entire procedures THA found significantly less proximal femoral stress
autonomously. For example, in this system, the preoperative shielding in the robotic group at 2 and 5 years [47]. This
plan is created using the ORTHODOC software, which cre- finding has been corroborated by a study of women 24 months
ates a 3D virtual model of the patient’s anatomy based on after robotic or conventional THA, in which proximal medial
preoperative CT. This plan is then transferred to the femoral spot welding was more prevalent (48%, 15/31 vs
ROBODOC surgical assistant, which completes all reaming, 11%, 3/27) and stress shielding was less prevalent (17%,
broaching, and positioning of the final femoral implant. The 5/31 vs 31%, 8/27) among ROBODOC patients [52].
surgeon oversees the robot and can activate an emergency stop Nakamura et al. found no significant difference in average
button, but does not directly control the robot. limb length inequality, but the ROBODOC group had signif-
Alternatively, “semi-active” systems offer “active con- icantly less variance in limb length inequality than the con-
straint,” in which the surgeon has ultimate control over the ventional THA group [47]. More patients with robotic THAs
surgical process, but receives auditory or tactile feedback from developed heterotopic ossification (27 vs 16%), although the
the robot to constrain the surgeon to a boundary defined by the difference was not significant.
384 Curr Rev Musculoskelet Med (2017) 10:378–387

Fig. 5 Mako RIO Robotic-

Assisted Total Hip Replacement
(Courtesy of Brad Waddell MD)

A primary benefit of navigation for THA is a reduction in Nakamura et al. reported a 2-point greater improvement
acetabular component positioning outliers. Acetabular cups (out of 100 points) in Japanese Orthopedic Association
implanted outside the “safe zone” identified by Lewinnek, (JOA) clinical scores for the robotic group at 2 and 3 years,
which is 5–25° of anteversion and 30–50° of abduction, are JOA scores did not differ at 5 years [47]. In a prospective
at increased risk for instability and dislocation [5•]. randomized controlled trial (RCT) comparing ROBODOC
Malpositioned cups, particularly those implanted in excessive and conventional THA through an anterolateral approach,
abduction, have also been linked to accelerated polyethylene Honl et al. reported better Mayo clinical and Harris hip scores
wear [53, 54]. Yet, multiple large retrospective studies have at 12 months, but no difference by 24 months [49]. In the same
demonstrated that with conventional free-hand techniques, study, robotic THA was associated with higher dislocation
only 60–85% of cups are implanted within this acceptable (18%, 11/61) and revision (15%, 3/80) rates, which the au-
window [55–57]. In a cadaver study, acetabular components thors attribute to abductor damage during robotic milling [49].
were placed more accurately with the Mako RIO system than Others found that no functional gait impairment when the
by conventional manual reaming [58]. These results are cor- abductor musculature was protected during ROBODOC sur-
roborated by clinical case series. Domb et al. found that 100% geries. Other potential technical complications include bone
(50/50) of robotic THA cups were within the Lewinnek safe motion during registration or cutting, which requires reregis-
zone, compared to only 80% (40/50) of conventionally placed tration, non-displaced femoral shaft fractures requiring
cups [59]. Similarly, Elson et al. report that 95% of cups were cerclage wiring, and milling defects in the acetabulum and
placed within 3.5° of the intended position [45]. Although few greater trochanter [61, 62].
studies have directly compared computer-navigated non-ro-
botic THA to robotic THA, one retrospective review of nearly
2000 THAs showed that both navigated and robotically Cost/Value Total Joint Arthroplasty
placed cups were significantly more likely to be within the
safe zone than conventional THA, and neither navigated nor The introduction of computer navigation and robotic-assisted
robotic THA was superior to the other [60]. arthroplasty has been shown to improve precision, improve
Femoral component size and positioning also contribute to patient satisfaction scores and lower blood loss, while lower-
the combined anteversion and leg length, which in turn affect ing certain complication rates. Their widespread acceptance,
hip stability, gait mechanics, and patient satisfaction. Robotic however, will ultimately depend on the cost-effectiveness and
THA has been shown to achieve limb length equality [47]. value added to joint replacement. Few studies have directly
Despite evidence supporting improved accuracy and fewer addressed this topic in relation to joint replacement. Moschetti
outliers with robotic THA, it is not yet clear whether these and colleagues performed a Markov decision analysis to de-
radiographic benefits translate into improved clinical out- termine the costs, outcomes, and incremental cost-
comes. Although some authors have reported better clinical effectiveness of robot-assisted UKA [63]. They created a
scores among patients with robotic THA at short-term follow- Markov decision model using a low-demand patient popula-
up, these improvements do not persist over longer follow-up tion at an age of 65 years. Their model found that case volume
periods. Harris hip scores in the original ROBODOC trial has the greatest influence on the cost-effectiveness of the ro-
were no different between groups at 2 years [48] or at 5– botic assistance, with a minimum of 94 cases per year (and a
7 years post-op [38]. revision rate below 1.2%) to become cost-effective over
Curr Rev Musculoskelet Med (2017) 10:378–387 385

conventional UKA. Further, at this setting, robotic UKA had a Compliance with Ethical Standards
slightly better outcome with 0.06 quality-adjusted life-years
Conflict of Interest Seth Jerabek reports consultancy fees from
per case. In another Markov decision model, Novak and col-
Stryker, outside of the submitted work. The other authors declare that
leagues estimated the cost-effectiveness of computer-assisted they have no conflict of interest.
navigation in total knee arthroplasty using known costs of the
technology and procedures and the known outcomes from Human and Animal Rights and Informed Consent This article does
review of the published literature [64]. They found that com- not contain any studies with human or animal subjects performed by any
puter assistance can be cost-effective if the cost of the tech- of the authors.
nology remained below $629 per case. This decision analysis
used current executed revision rates at 15 years and currently
published rates of achieving coronal plane alignment within References
3° of neutral. Using a Markov model as well, Slover and co-
authors found that computer assistance becomes less cost- Papers of particular interest, published recently, have been
effective as the volume of cases decreases and as the cost of highlighted as:
the technology increases [65]. Further, they found that revi- • Of importance
sion rate must decrease for the technology to be cost-effective.
In higher volume centers, the 20-year revision rate would have 1. Robertsson O, Dunbar M, Pehrsson T, Knutson K, Lidgren L.
to decrease by 2% to maintain cost-effectiveness and in lower Patient satisfaction after knee arthroplasty: a report on 27,372 knees
operated on between 1981 and 1995 in Sweden. Acta Orthop
volume centers (25 arthroplasties per year), the revision rate Scand. 2000;71:262–7. doi:10.1080/000164700317411852.
would have to decrease by 13% to become cost-effective. To 2. Warth LC, Callaghan JJ, Liu SS, Klaassen AL, Goetz DD, Johnston
the point of reducing revision rates, de Steiger and colleagues RC. Thirty-five-year results after Charnley total hip arthroplasty in
did show a reduction in revision rate in the Australian joint patients less than fifty years old. A concise follow-up of previous
registry, as mentioned previously [16]. reports. J Bone Joint Surg Am. 2014;96:1814–9. doi:10.2106/JBJS.
M.01573.
Not all studies show that technologies in arthroplasty are 3. Knight SR, Aujla R, Biswas SP. Total hip arthroplasty - over 100
cost-effective. Burnett and Barrack discuss the cost and com- years of operative history. Orthop Rev. 2011;3:e16. doi:10.4081/or.
plications associated with total knee artrhoplasty computer 2011.e16.
navigation [19]. They note the increased surgical time, poten- 4.• Berend ME, Ritter MA, Meding JB, Faris PM, Keating EM,
Redelman R, et al. Tibial component failure mechanisms in total
tial need for costly preoperative imaging, the cost of the tech-
knee arthroplasty. Clin Orthop. 2004;428:26–34. This study de-
nology and disposables, and the potential complications tails failure mechanisms among 3152 metal backed cemented
unique only to these technologies (pin site infection and tibial components in total knee arthroplasty.Overall tibial revi-
fracture). sion rate was 1.3% (41 tobial components). Primary mode of
failure was medial bone collapse and main factors related to
this were tibial component varus greater than 3 degrees, higher
BMI and overall postoperative varun alignment of the limb.
Summary 5.• Lewinnek GE, Lewis JL, Tarr R, Compere CL, Zimmerman JR.
Dislocations after total hip-replacement arthroplasties. J Bone Joint
In regards to the value added to arthroplasty, there is currently Surg Am. 1978;60:217–20. This classic study defined the safe
zones for acetabular cup position in total hip arthroplasty.
no definitive answer as to how much, or if any, value is added Safe zones were defined as inclination of 40 +/- 10 degrees and
to joint arthroplasty with these newer technologies. Many anterversion of 15 +/-10 degrees. Outside these zones, the dis-
studies have shown improved outcomes, improved accuracy, location rate increased from 1.5% to 6.1%.
and lower revision rates using these technologies. However, 6. Wasielewski RC, Galante JO, Leighty RM, Natarajan RN,
Rosenberg AG. Wear patterns on retrieved polyethylene tibial in-
the question remains if the extra cost associated with these
serts and their relationship to technical considerations during total
technologies will keep them viable. knee arthroplasty. Clin Orthop. 1994;299:31–43.
There are many potential benefits to CAS for the future. 7. Jeffery RS, Morris RW, Denham RA. Coronal alignment after total
Beal et al. describe unknown variables that we may be able to knee replacement. J Bone Joint Surg Br. 1991;73:709–14.
elucidate as we use these technologies to solve current prob- 8. Barrack RL, Schrader T, Bertot AJ, Wolfe MW, Myers L.
Component rotation and anterior knee pain after total knee
lems. Future areas of interest can include teaching, research,
arthroplasty. Clin Orthop. 2001;392:46–55.
rotational alignment of the femur, and surgical documentation 9. Brown MJ, Matthews JR, Bayers-Thering MT, Phillips MJ,
opportunities [66]. Krackow KA. Low incidence of postoperative complications with
Better precision and accuracy should relate to longer sur- navigated Total knee arthroplasty. J Arthroplast. 2017; doi:10.1016/
vival and better outcomes. As technology becomes more j.arth.2017.01.045.
10. Knee Navigation Application for Arthroplasty Knee3 from
mainstream, hopefully, it will become easier and faster to Brainlab. Brainlab n.d. https://www.brainlab.com/en/surgery-
use and at a more neutral cost. It is the authors’ opinion that products/orthopedic-surgery-products/knee-navigation-
these technologies are a vital part of the future of orthopaedics. application/. Accessed 1 April 2017.
386 Curr Rev Musculoskelet Med (2017) 10:378–387

11. Computer-Assisted Surgery — Exactech, Inc. n.d. https://www. 27. Bell SW, Anthony I, Jones B, MacLean A, Rowe P, Blyth M.
exac.com/products/knee/advanced-surgical-instrumentation. Improved accuracy of component positioning with robotic-
Accessed 1 April 2017. assisted Unicompartmental knee arthroplasty: data from a prospec-
12. Nam D, Jerabek SA, Haughom B, Cross MB, Reinhardt KR, tive, randomized controlled study. J Bone Joint Surg Am. 2016;98:
Mayman DJ. Radiographic analysis of a hand-held surgical navi- 627–35. doi:10.2106/JBJS.15.00664.
gation system for tibial resection in total knee arthroplasty. J 28. Pearle AD, van der List JP, Lee L, Coon TM, Borus TA, Roche
Arthroplast. 2011;26:1527–33. doi:10.1016/j.arth.2011.01.012. MW. Survivorship and patient satisfaction of robotic-assisted me-
13.• Cip J, Widemschek M, Luegmair M, Sheinkop MB, Benesch T, dial unicompartmental knee arthroplasty at a minimum two-year
Martin A. Conventional versus computer-assisted technique for to- follow-up. Knee. 2017;24:419–28. doi:10.1016/j.knee.2016.12.
tal knee arthroplasty: a minimum of 5-year follow-up of 200 pa- 001.
tients in a prospective randomized comparative trial. J Arthroplast. 29. Lonner JH, Smith JR, Picard F, Hamlin B, Rowe PJ, Riches PE.
2014;29:1795–802. doi:10.1016/j.arth.2014.04.037. This study High degree of accuracy of a novel image-free handheld robot for
compared 100 consecutive navigated total knee replacements unicondylar knee arthroplasty in a cadaveric study. Clin Orthop.
to 100 consecutive conventional total knee replacements. At 5 2015;473:206–12. doi:10.1007/s11999-014-3764-x.
years postop, they found the navigated group to have a non- 30. Nam D, Weeks KD, Reinhardt KR, Nawabi DH, Cross MB,
statistically lower rate of revision (1.1% vs 4.6%, p=0.368). The Mayman DJ. Accelerometer-based, portable navigation vs
navigated group had a higher chance of being within 3 degrees imageless, large-console computer-assisted navigation in total knee
of the mechanical axis and more accurate slope. Finally, Insall arthroplasty: a comparison of radiographic results. J Arthroplast.
and HSS scores were higher in the navigated group. 2013;28:255–61. doi:10.1016/j.arth.2012.04.023.
14. Blakeney WG, Khan RJK, Palmer JL. Functional outcomes follow- 31. Steppacher SD, Kowal JH, Murphy SB. Improving cup positioning
ing total knee arthroplasty: a randomised trial comparing computer- using a mechanical navigation instrument. Clin Orthop. 2011;469:
assisted surgery with conventional techniques. Knee. 2014;21:364– 423–8. doi:10.1007/s11999-010-1553-8.
8. doi:10.1016/j.knee.2013.04.001. 32. Dorr LD, Malik A, Wan Z, Long WT, Harris M. Precision and bias
15. Baumbach JA, Willburger R, Haaker R, Dittrich M, Kohler S. 10- of imageless computer navigation and surgeon estimates for acetab-
year survival of navigated versus conventional TKAs: a retrospec- ular component position. Clin Orthop. 2007;465:92–9. doi:10.
tive study. Orthopedics. 2016;39:S72–6. doi:10.3928/01477447- 1097/BLO.0b013e3181560c51.
20160509-21. 33. Davis ET, Schubert M, Wegner M, Haimerl M. A new method of
16. de Steiger RN, Liu Y-L, Graves SE. Computer navigation for total registration in navigated hip arthroplasty without the need to regis-
knee arthroplasty reduces revision rate for patients less than sixty- ter the anterior pelvic plane. J Arthroplast. 2015;30:55–60. doi:10.
five years of age. J Bone Joint Surg Am. 2015;97:635–42. doi:10. 1016/j.arth.2014.08.026.
2106/JBJS.M.01496. 34.• Lass R, Kubista B, Olischar B, Frantal S, Windhager R, Giurea A.
17. Licini DJ, Meneghini RM. Modern abbreviated computer naviga- Total hip arthroplasty using imageless computer-assisted hip navi-
tion of the femur reduces blood loss in total knee arthroplasty. J gation: a prospective randomized study. J Arthroplast. 2014;29:
Arthroplast. 2015;30:1729–32. doi:10.1016/j.arth.2015.04.020. 786–91. doi:10.1016/j.arth.2013.08.020. In this randomized
18. Moskal JT, Capps SG, Mann JW, Scanelli JA. Navigated versus study, the authors found that computer navigation did not aid
conventional total knee arthroplasty. J Knee Surg. 2014;27:235– in improving the inclination, but did find the anteversion to be
48. doi:10.1055/s-0033-1360659. significantly more accurate. They also found outliers from the
19. Burnett RSJ, Barrack RL. Computer-assisted total knee arthroplasty safe zone to be significantly less in the navigated group.
is currently of no proven clinical benefit: a systematic review. Clin 35. Weber M, Woerner M, Springorum R, Sendtner E, Hapfelmeier A,
Orthop. 2013;471:264–76. doi:10.1007/s11999-012-2528-8. Grifka J, et al. Fluoroscopy and imageless navigation enable an
20. Song EK, Agrawal PR, Kim SK, Seo HY, Seon JK. A randomized equivalent reconstruction of leg length and global and femoral off-
controlled clinical and radiological trial about outcomes of set in THA. Clin Orthop. 2014;472:3150–8. doi:10.1007/s11999-
navigation-assisted TKA compared to conventional TKA: long- 014-3740-5.
term follow-up. Knee Surg Sports Traumatol Arthrosc Off J 36. Ellapparadja P, Mahajan V, Deakin AH, Deep K. Reproduction of
ESSKA. 2016;24:3381–6. doi:10.1007/s00167-016-3996-2. hip offset and leg length in navigated Total hip arthroplasty: how
21. Song E-K, Seon J-K, Park S-J, Jung WB, Park H-W, Lee GW. accurate are we? J Arthroplast. 2015;30:1002–7. doi:10.1016/j.arth.
Simultaneous bilateral total knee arthroplasty with robotic and con- 2015.01.027.
ventional techniques: a prospective, randomized study. Knee Surg 37. Renkawitz T, Weber M, Springorum H-R, Sendtner E, Woerner M,
Sports Traumatol Arthrosc Off J ESSKA. 2011;19:1069–76. doi: Ulm K, et al. Impingement-free range of movement, acetabular
10.1007/s00167-011-1400-9. component cover and early clinical results comparing “femur-first”
22. Ponder CE, Plaskos C, Cheal EJ. Press-fit Total knee arthroplasty navigation and “conventional” minimally invasive total hip
with a robotic-cutting guide: proof of concept and initial clinical arthroplasty: a randomised controlled trial. Bone Jt J. 2015;97–B:
experience. Bone Jt J. 2013;95–B:61. 890–8. doi:10.1302/0301-620X.97B7.34729.
23. Koenig JA, Suero EM, Plaskos C. Surgical accuracy and efficiency 38. Keshmiri A, Schröter C, Weber M, Craiovan B, Grifka J,
of computer-navigated Tka with a robotic cutting guide – report on Renkawitz T. No difference in clinical outcome, bone density and
the first 100 cases. Orthop Proc. 2012;94–B:103. polyethylene wear 5-7 years after standard navigated vs. conven-
24. Hampp E, Scholl L, Prieto M, Chang T, Abbasi A, Stoker M, et al. tional cementfree total hip arthroplasty. Arch Orthop Trauma Surg.
Robotic-arm assisted total knee arthroplasty demonstrated greater 2015;135:723–30. doi:10.1007/s00402-015-2201-2.
accuracy to plan compared to manual technique. 2017. 39. Stiehl JB, Heck DA, Jaramaz B, Amiot L-P. Comparison of fluoro-
25. Hampp E, Stoker M, Scholl L, Otto J, Mont M. Robotic-arm scopic and imageless registration in surgical navigation of the ace-
assisted total knee arthroplasty demonstrated soft tissue protection. tabular component. Comput Aided Surg Off J Int Soc Comput
2017. Aided Surg. 2007;12:116–24. doi:10.3109/10929080701292939.
26. Riddle DL, Jiranek WA, McGlynn FJ. Yearly incidence of 40. Kalteis T, Handel M, Bäthis H, Perlick L, Tingart M, Grifka J.
unicompartmental knee arthroplasty in the United States. J Imageless navigation for insertion of the acetabular component in
Arthroplast. 2008;23:408–12. doi:10.1016/j.arth.2007.04.012. total hip arthroplasty: is it as accurate as CT-based navigation? J
Curr Rev Musculoskelet Med (2017) 10:378–387 387

Bone Joint Surg Br. 2006;88:163–7. doi:10.1302/0301-620X. 54. Leslie IJ, Williams S, Isaac G, Ingham E, Fisher J. High cup angle
88B2.17163. and microseparation increase the wear of hip surface replacements.
41. Digioia AM, Jaramaz B, Plakseychuk AY, Moody JE, Nikou C, Clin Orthop. 2009;467:2259–65. doi:10.1007/s11999-009-0830-x.
Labarca RS, et al. Comparison of a mechanical acetabular align- 55. Callanan MC, Jarrett B, Bragdon CR, Zurakowski D, Rubash HE,
ment guide with computer placement of the socket. J Arthroplast. Freiberg AA, et al. The John Charnley award: risk factors for cup
2002;17:359–64. malpositioning: quality improvement through a joint registry at a
42.• Parratte S, Argenson J-NA. Validation and usefulness of a tertiary hospital. Clin Orthop. 2011;469:319–29. doi:10.1007/
computer-assisted cup-positioning system in total hip arthroplasty. s11999-010-1487-1.
A prospective, randomized, controlled study. J Bone Joint Surg 56. Bosker BH, Verheyen CCPM, Horstmann WG, Tulp NJA. Poor
Am. 2007;89:494–9. doi:10.2106/JBJS.F.00529. In this study, accuracy of freehand cup positioning during total hip arthroplasty.
computer navigation took longer, however, it gave a Arch Orthop Trauma Surg. 2007;127:375–9. doi:10.1007/s00402-
significant reduction in the outliers from the safe zone for 007-0294-y.
inclination and version. 57. Leichtle U, Gosselke N, Wirth CJ, Rudert M. Radiologic evaluation
43. Sugano N, Takao M, Sakai T, Nishii T, Miki H. Does CT-based of cup placement variation in conventional total hip arthroplasty.
navigation improve the long-term survival in ceramic-on-ceramic ROFO Fortschr Geb Rontgenstr Nuklearmed. 2007;179:46–52.
THA? Clin Orthop. 2012;470:3054–9. doi:10.1007/s11999-012- doi:10.1055/s-2006-927085.
2378-4. 58. Nawabi DH, Conditt MA, Ranawat AS, Dunbar NJ, Jones J, Banks
44. Beckmann J, Stengel D, Tingart M, Götz J, Grifka J, Lüring C. S, et al. Haptically guided robotic technology in total hip
Navigated cup implantation in hip arthroplasty. Acta Orthop. arthroplasty: a cadaveric investigation. Proc Inst Mech Eng [H].
2009;80:538–44. doi:10.3109/17453670903350073. 2013;227:302–9. doi:10.1177/0954411912468540.
59. Domb BG, El Bitar YF, Sadik AY, Stake CE, Botser IB.
45. Elson L, Dounchis J, Illgen R, Marchand RC, Padgett DE, Bragdon
Comparison of robotic-assisted and conventional acetabular cup
CR, et al. Precision of acetabular cup placement in robotic integrat-
placement in THA: a matched-pair controlled study. Clin Orthop.
ed total hip arthroplasty. Hip Int J Clin Exp Res Hip Pathol Ther.
2014;472:329–36. doi:10.1007/s11999-013-3253-7.
2015;25:531–6. doi:10.5301/hipint.5000289.
60. Domb BG, Redmond JM, Louis SS, Alden KJ, Daley RJ, LaReau
46. Bargar WL. Robots in orthopaedic surgery: past, present, and fu- JM, et al. Accuracy of component positioning in 1980 Total hip
ture. Clin Orthop. 2007;463:31–6. arthroplasties: a comparative analysis by surgical technique and
47. Nakamura N, Sugano N, Nishii T, Kakimoto A, Miki H. A com- mode of guidance. J Arthroplast. 2015;30:2208–18. doi:10.1016/j.
parison between robotic-assisted and manual implantation of arth.2015.06.059.
Cementless Total hip arthroplasty. Clin Orthop. 2010;468:1072– 61. Schulz AP, Seide K, Queitsch C, von Haugwitz A, Meiners J,
81. doi:10.1007/s11999-009-1158-2. Kienast B, et al. Results of total hip replacement using the
48. Bargar WL, Bauer A, Börner M. Primary and revision total hip Robodoc surgical assistant system: clinical outcome and evaluation
replacement using the Robodoc system. Clin Orthop. 1998:82–91. of complications for 97 procedures. Int J Med Robot Comput Assist
49. Honl M, Dierk O, Gauck C, Carrero V, Lampe F, Dries S, et al. Surg MRCAS. 2007;3:301–6. doi:10.1002/rcs.161.
Comparison of robotic-assisted and manual implantation of a pri- 62. Chun YS, Kim KI, Cho YJ, Kim YH, Yoo MC, Rhyu KH. Causes
mary total hip replacement. A prospective study. J Bone Joint Surg and patterns of aborting a robot-assisted arthroplasty. J Arthroplast.
Am. 2003;85–A:1470–8. 2011;26:621–5. doi:10.1016/j.arth.2010.05.017.
50. Spencer EH. The ROBODOC clinical trial: a robotic assistant for 63. Moschetti WE, Konopka JF, Rubash HE, Genuario JW. Can robot-
total hip arthroplasty. Orthop Nurs. 1996;15:9–14. assisted Unicompartmental knee arthroplasty be cost-effective?
51. Nishihara S, Sugano N, Nishii T, Miki H, Nakamura N, Yoshikawa A Markov decision analysis. J Arthroplast. 2016;31:759–65. doi:
H. Comparison between hand rasping and robotic milling for stem 10.1016/j.arth.2015.10.018.
implantation in cementless total hip arthroplasty. J Arthroplast. 64. Novak EJ, Silverstein MD, Bozic KJ. The cost-effectiveness of
2006;21:957–66. doi:10.1016/j.arth.2006.01.001. computer-assisted navigation in total knee arthroplasty. J Bone
52. Hananouchi T, Sugano N, Nishii T, Nakamura N, Miki H, Joint Surg Am. 2007;89:2389–97. doi:10.2106/JBJS.F.01109.
Kakimoto A, et al. Effect of robotic milling on periprosthetic bone 65. Slover JD, Tosteson ANA, Bozic KJ, Rubash HE, Malchau H.
remodeling. J Orthop Res Off Publ Orthop Res Soc. 2007;25: Impact of hospital volume on the economic value of computer
1062–9. doi:10.1002/jor.20376. navigation for total knee replacement. J Bone Joint Surg Am.
53. Gallo J, Havranek V, Zapletalova J. Risk factors for accelerated 2008;90:1492–500. doi:10.2106/JBJS.G.00888.
polyethylene wear and osteolysis in ABG I total hip arthroplasty. 66. Beal MD, Delagramaticas D, Fitz D. Improving outcomes in total
Int Orthop. 2010;34:19–26. doi:10.1007/s00264-009-0731-3. knee arthroplasty-do navigation or customized implants have a
role? J Orthop Surg. 2016;11:60. doi:10.1186/s13018-016-0396-8.