Blue Cross and Blue Shield of Montana (BCBSMT) considers computer-assisted surgery for orthopedic procedure of the pelvis and appendicular skeleton experimental, investigational and unproven.
Trauma or Fracture
Computer-assisted surgery has been most frequently mentioned as an adjunct to pelvic, acetabular, or femoral fractures. For example, fixation of these fractures typically requires percutaneous placement of screws or guide wires. Conventional fluoroscopic guidance (i.e., C-arm fluoroscopy) provides imaging in only one plane. Therefore, the surgeon must position the implant in one plane, and then get additional images in other planes in a trial and error fashion to ensure that the device has been properly placed. This process adds significant time in the operating room (OR) and radiation exposure. It is hoped the computer-assisted surgery would allow for minimally invasive fixation and provide more versatile screw trajectories with less radiation exposure. Therefore, computed-assisted surgery is considered an alternative to the existing image guidance using C-arm fluoroscopy.
Ideally, one would like controlled trials comparing OR time, radiation exposure, and long-term outcomes of those whose surgery was conventionally guided using C-arm versus image-guided using computer-assisted surgery. While several in vitro and review studies have been published, a literature search identified one clinical trial of computer-assisted surgery in trauma or fracture cases. Suhm and colleagues reported on a case series of 27 patients with femoral fractures who underwent implantation of a femoral nail. Outcomes included precision of interlocking, exposure time, and OR time. Without a control or comparison group, it is not possible to determine the impact of the computer assistance.
Arthroplasty (total hip [THA] and total knee [TKA])
For both total hip and knee arthroplasties, optimal alignment is considered an important aspect of long-term success. Malalignment of arthroplasty components is one of the leading causes of instability and reoperation. In THA, orientation of the acetabular component of the THA is considered critical, while for TKA, alignment of the femoral and tibial components as well as ligament balancing are considered important outcomes. The alignment of the knee prosthesis can be measured along several different axes, including the mechanical axis, and the frontal and sagittal axes of both the femur and tibia. It is proposed that computer-assisted surgery improves the alignments of the various components of THA and TKA. Ideally, one would like controlled trials comparing the long-term outcomes, including stability and reoperation rates. Intermediate outcomes include the percentage of implants that achieve a predetermined level of acceptable alignment.
Total Hip Arthroplasty (THA)
No controlled trials regarding total hip arthroplasty are available. In an uncontrolled case series, Leenders and colleagues studied the variability in placement of the acetabular component among three groups of patients: 1) those undergoing THA using free-hand placement before computer-assisted surgery was available; 2) those undergoing THA with computer assistance; and 3) those undergoing free-hand placement after computer assistance was available. While variability was reduced between groups one and two, there was no significant difference between groups two and three. No data regarding long-term outcome was reported. Digioia and colleagues reported on a case series of 78 patients (82 hips) who underwent THA and compared the alignment directed by a mechanical guide and computer assistance. The authors hypothesized that the use of the mechanical guide rather than computer assistance would have resulted in an unacceptable acetabular alignment in 78% of hips. In summary, data are inadequate to permit scientific conclusions regarding computer-assisted surgery for THA.
Total Knee Arthroplasty (TKA)
Five randomized trials have focused on computer-assisted TKA. Saragaglia and colleagues randomized 25 patients to receive computer-assisted TKA and 25 to conventional TKA. The principal outcome of the procedure was the achievement of target alignment of the prosthesis. There was no significant difference in outcomes between the two groups; both met the target orientation of mechanical alignment of 0–3 degrees. Decking and colleagues reported on a similarly designed study of 52 patients randomized to computer-assisted or conventional TKA. The primary outcome was alignment measured three months postoperatively. While both groups showed a tendency for varus or valgus deviation of the mechanical axis of the leg, eight of the manually implanted knees versus only one computed-assisted implanted knee showed a deviation of five degrees or more, a statistically significant difference. Other radiologic measures of alignment were not significantly different between the two groups. Stockl and colleagues conducted a trial randomizing 64 patients to undergo computer-assisted navigation or conventional TKA. A variety of measures of alignment were improved in the navigated group. Sparmann and colleagues reported on the largest study of 240 patients randomized to replacement with or without computer- assisted navigation. A total of 97.5 % in the navigated group had a mechanical alignment between 0–2 degrees, compared to 77.5% in the conventional group. Finally, Victor and Hoste studied 100 patients who were randomized to conventional or image-guided computer navigated TKA. In the navigated group all patients showed alignment of the mechanical axis between 0 and 2 degrees. In contrast, in the conventional group, only 73.2% achieved mechanical alignment between 0 and 2 degrees.
Several case series have also been published. Haaker and colleagues published the largest case series of 100 TKAs inserted with an imageless computer navigation system, compared with a matched control group of conventionally implanted knees. An excellent outcome was defined as 0–3 degrees of deviation on the mechanical axis and 0–2 degrees of deviation on other axes of alignment. The percentage of excellent results was significantly higher in the navigated groups for all alignment measures except for the sagittal tibial axis angle. Zorman and Etuin reported on the axis alignment of 72 TKAs performed with navigational assistance compared to a historical cohort of 62 TKAs performed with conventional instrumentation. There was a highly significant improvement in the alignment along the mechanical axis in the navigated group; all of those in the navigated group showed neutral alignment, while 47% of those in the conventional group showed a deviation of the mechanical axis of more than two degrees from neutral alignment. Jenny and Boeri compared the outcomes of 30 patients undergoing computer-assisted TKA with 30 matched and paired patients. The outcome studied was the femorotibial angle. The authors concluded that those in the computer-assisted group showed an improved quality of implantation. Other case series have also reported improved alignment using computer-assisted navigation systems.
In summary, four randomized studies and several case series have consistently shown that computer-assisted navigation is associated with improved postoperative alignment along several different axes. However, the published studies report only immediate postoperative outcomes, and there are no reports of patient-oriented outcomes, such as pain, range of motion, or reoperation rate. Thus there are inadequate scientific data to permit conclusions regarding whether the improvement in alignment associated with computer-assisted navigation will result in significant clinical improvement in patients undergoing TKA.
Trauma or Fracture
Several studies reported the use of CAN for ACL reconstruction. One of the studies randomized 60 patients to either standard instrumentation or computer-assisted guidance for tunnel placement with follow-up at 1, 3, 6, 12, 18 and 24 months. There were no differences between the groups in measurements of laxity. However, there was less variability in side-to-side anterior laxity in the navigated group (e.g., 97% were within 2 mm of laxity in the navigated group vs. 83% in the conventional group at an applied force of 150 Newtons). There was a significant difference in the sagittal position of the tibial tunnel (distance from the Blumensaat line of 0.4 vs. -1.2 mm), suggesting possible impingement in extension for the conventional group. At the final follow-up (24 months) all knees had normal function, with no differences observed between the groups.
CAN for internal fixation of femoral neck fractures was described in a retrospective analysis consisting of two cohorts of consecutive patients (20 each, performed from 2001 to 2003 at two different campuses of a medical center) who underwent internal fixation with three screws for a femoral neck fracture. Three of five measurements of parallelism and neck coverage were significantly improved by CAN; these included a larger relative neck area held by the screws (32% vs. 23%) and less deviation on the lateral projection for both the shaft (1.7 vs. 5.2 degrees) and the fracture (1.7 vs. 5.5 degrees) screw angles. Slight improvements in anteroposterior screw angles (1.3 vs. 2.1 and 1.3 vs. 2.4 degrees) did not reach statistical significance. There were two reoperations in the CAN group and six in the conventional group. Complications (collapse, subtrochanteric fracture, head penetration, osteonecrosis) were lower in the CAN group (3 vs. 11). Additional controlled studies are needed.
Total Knee Arthroplasty
A recent TEC Assessment evaluated computer-assisted navigation for total knee replacement.) Nine randomized controlled trials (RCTs) were reviewed. Criteria for the RCTs included having at least 25 patients per group and comparing limb alignment, surgical or functional outcomes following TKA with CAN or conventional methods. Also reviewed were cohort and case series that evaluated long-term associations between malalignment of prosthetic components and poor outcomes. In the largest of the cohort studies, which included over 2,000 patients (3,000 knees) with an average of five-year follow-up, 41 revisions for tibial component failure (1.3% of the cohort) were identified. The risk ratio for age was estimated at 8.3, with a greater risk observed in younger, more active patients. For malalignment (defined as > 3 degrees varus or valgus), the risk ratio was estimated to be 17.3.
The combined data from the prospective RCTs showed:
- A significant decrease in the percentage of limbs considered to be outliers (e.g., > 3 degrees of varus or valgus from a neutral mechanical axis) with CAN. In the conventional group, 33% of patients had malalignment of the overall femoral/tibial axis. In the navigated group, 18% of patients were considered to have malalignment of the mechanical axis. For the combined data set there was a decrease in malalignment in 15% of patients, with an estimated number needed to treat (NNT) of 6.7 to avoid one case of malalignment.
- Surgical time increased by 10 to 20 minutes in all but one study. CAN-associated reduction in blood loss was less consistent, with only some of the studies showing a decrease in blood loss of 100 to 200 mls.
- RCTs that assessed function (up to 2 years follow-up) did not find evidence of improved health outcomes. However, the studies were not adequately powered to detect functional differences, and data on long-term follow-up is not available.
The report concluded that there is currently no direct evidence to support an improvement in clinical outcomes with CAN for TKA. As a result of deficiencies in the available evidence (e.g., potential for bias in observational studies and lack of long-term follow-up in the RCTs), it is not possible to determine whether the degree of improvement in alignment that has been reported in the RCTs leads to meaningful improvements in clinically relevant outcomes such as pain, function or revision surgery.
A meta-analysis of CAN for TKA was conducted that included 33 studies and 3,423 patients. The studies were of varying methodological quality and included eleven randomized trials. Although no significant difference in mechanical axes between the navigated and conventional surgery group was found, navigated surgery was found to result in a lower risk of malalignment. It was calculated that one of every five patients would avoid unfavorable component positioning (greater than 3 degrees) with CAS. The authors concluded that methodological weaknesses of the available trials limited the conclusions of the meta-analysis and no conclusive inferences could be reached for functional outcomes or complication rates.
Consistent with previous conclusions, the literature supports a decrease in variability of joint alignment with computer-assisted TKA, particularly with respect to the number of outliers. Although some observational data suggest that malalignment may increase the probability of early failure, recent RCTs do not show improved health outcomes with CAN. Given the low short-term revision rates associated with conventional TKA procedures and the inadequate power of available studies to detect changes in function, studies that assess health outcomes in a larger number of subjects with longer follow-up are needed.
Cup Positioning in Total Hip Arthroplasty and Periacetabular Osteotomy:
One recent study randomized patients to CAN for THA (n=30) or freehand cup positioning (n=30) by an experienced surgeon. The mean additional time for the computer-assisted procedure was twelve minutes. There was no difference between the computer-assisted group and the freehand-placement group with regard to the mean abduction or anteversion angles measured by CT. A smaller variation in the positioning of the acetablular component was observed in the CAN group; 20% of cup placements were considered to be outliers in the CAN group compared with 57% in the freehand-placement group. Another study randomly assigned 36 patients with symptomatic adult dysplastic hip to either CT-based navigation or the conventional technique for periacetabular osteotomy. An average of 0.6 intraoperative radiographs were taken in the navigated group compared with 4.4 in the conventional group, resulting in a total operative time that was 21 minutes shorter for CAN. There were no differences between the groups for correction in femoral head coverage or for functional outcomes (pain, walking, and range of motion) at 24 months.
A search and review of scientific literature conducted through October 2007 did not identify any published peer-reviewed literature that addresses the limitations noted in the above discussion. Therefore, the coverage position of this policy remains investigational.
A search of peer reviewed literature through November 2010. In 2008, Luring and colleagues published results from a 3-arm randomized trial (30 patients per group) that compared minimally-invasive TKA, with or without computer-assisted navigation, and conventional TKA. In this study the mini-incision averaged 13 cm (range: 10-14 cm) while the conventional midline incision averaged 17 cm (range: 15-19 cm); both were performed with a medial parapatellar approach. In addition, with the minimally-invasive procedure there was subluxation rather than eversion of the patella and no tibio-femoral dislocation. Postoperative rehabilitation and hospital stay were not described. On average, the surgical procedure took longer in the computer-assisted minimally invasive surgery (MIS) group (58 min) compared to the conventional (44 min) and free-hand MIS group (40 min), and was associated with greater blood loss. Independent evaluation of postoperative radiographs showed reduced deviation in mechanical axis alignment in the computer-assisted navigation group (1.0 degree) compared to both the freehand minimally-invasive group (1.8 degrees) and the conventional TKA group (2.1 degrees). Compared to three outliers in the freehand minimally-invasive group and two outliers in the conventional TKA group, no outliers greater than three degrees were observed in the computer-assisted minimally invasive group. Follow-up (100%) with the Knee Society Score (KSS) and WOMAC at 1, 6, and 12 weeks revealed no differences between the three groups. Since there was no statistically significant clinical difference at six or twelve weeks, the planned six and twelve-month follow-up was stopped. The authors concluded that according to patient satisfaction (WOMAC) and clinical outcome (KSS), the minimally invasive approach in TKA is still not proven.
Overall, the literature supports a decrease in variability of alignment with CAN, particularly with respect to the number of outliers. Although some observational data suggest that malalignment may increase the probability of early failure, recent RCTs with short to mid-term follow-up have not shown improved health outcomes with CAN. Given the low short-term revision rates associated with conventional procedures and the inadequate power of available studies to detect changes in function, studies that assess health outcomes in a larger number of subjects with longer follow-up are needed. The most promising utilization of this procedure appears to be the ability to decrease incision length without loss of accuracy in component alignment. Although evidence at this time has not adequately demonstrated improved health outcomes with this more resource-intensive combination, continued technology development in this area is expected.
Disclaimer for coding information on Medical Policies
Procedure and diagnosis codes on Medical Policy documents are included only as a general reference tool for each policy. They may not be all-inclusive.
The presence or absence of procedure, service, supply, device or diagnosis codes in a Medical Policy document has no relevance for determination of benefit coverage for members or reimbursement for providers. Only the written coverage position in a medical policy should be used for such determinations.
Benefit coverage determinations based on written Medical Policy coverage positions must include review of the member’s benefit contract or Summary Plan Description (SPD) for defined coverage vs. non-coverage, benefit exclusions, and benefit limitations such as dollar or duration caps.