Introduction
Substantial technological innovation has been made in spine surgery over the last few decades. This innovation has ranged from improvement in operative techniques, surgical implants, and biologics to improved accuracy with image-guided navigation and robotics. Robotic assistance is being increasingly utilized across multiple surgical specialties including urology, obstetrics and gynecology (OB/GYN), orthopedics, and neurosurgery. Although the utilization of robotics in spine surgery is still in its nascent stages, the recent literature has demonstrated that it has the potential to revolutionize safety and accuracy, not just for the placement of spinal instrumentation, but also to complete other critical operative steps while minimizing radiation exposure to the surgical team. However, while many providers and institutions are acquiring various surgical robots, there have been reports that the technology has failed to penetrate into routine practice. To demonstrate its value in healthcare, any new technology must demonstrate cost-effectiveness in addition to quality or improvement in outcomes. Spine providers have been hesitant to incorporate robotics and navigation into routine surgical care due to the notion of substantially higher costs associated with the acquisition of these newer guidance platforms. In today’s pay-for-performance era, spine surgeons are under increasing pressure and scrutiny to develop measures that control pricing and allow increased accountability for surgical performance. The incorporation of robotic and navigation systems into routine practice might allow implementation of such value-based spinal care.
In addition to quality and safety, the cost-effectiveness of robotic surgery has been evaluated within other surgical specialties such as urology and OB/GYN. The cumulative evidence has significantly changed practice patterns in urologic surgery. Such cost-effectiveness analyses are essential to obtain a more balanced perspective of practice utility and are imperative to design and effect solutions to improve performance. However, such analyses are limited in the realm of spine surgery. In this chapter, the authors discuss the available evidence on the cost-effectiveness of robotics and navigation in spine surgery and provide an overview of the mechanisms regarding how the assimilation of such platforms into the daily surgical workflow could provide long-term cost savings.
Scope of Robotic and Navigated Spine Surgery
A variety of image-guided navigation and robotic platforms are currently available in the spine surgery market. Although navigation and robotic guidance have historically been mutually exclusive, the newer platforms allow the integration of the two modalities to facilitate real-time feedback in addition to the placement of instrumentation along pre-planned trajectories with the use of a robotic arm. Each type of platform achieves stereotaxy through some form of preoperative or intraoperative radiographic imaging such as x-ray, MRI, computer tomography (CT) or three-dimensional (3D) fluoroscopy to generate a comprehensive spinal map for precision and accuracy. Intraoperative cone-beam CT or O-arm (“computer-assisted navigation”) and 3D fluoroscopy (“virtual fluoroscopy”) are the most commonly utilized imaging modalities in modern surgical navigation systems. Both allow frameless stereotaxy with real-time, navigated feedback of instruments such as tubular dilators, screwdrivers, drills, and awls. Examples of such systems include the Airo Mobile Intraoperative CT-based Spinal Navigation (Brainlab©, Feldkirchen, Germany), Stryker Spinal Navigation with SpineMask© Tracker, and SpineMap Software (Stryker©, Kalamazoo, Michigan), StealthStation Spine Surgery Imaging, and Surgical Navigation with O-arm (Medtronic©, Minneapolis, Minnesota), and Ziehm Vision FD Vario 3D with NaviPort integration (Ziehm Imaging©, Orlando, Florida). It is essential to note that instrument maneuvering with these systems is still entirely surgeon-dependent with the navigation system merely providing anatomic feedback.
Robotic systems allow user-operated trajectory planning using radiographic guidance, following by “locking” of the robotic arm along the desired screw trajectory. This can be additionally aided by intraoperative navigation to allow real-time feedback in more recent platforms. Surgical robots can be divided into three types depending on the levels of assistance: (1) supervisory controlled systems, in which the robot performs actions based on a pre-planned trajectory under close surgeon supervision; (2) tele-surgical systems (e.g., da Vinci robot, Intuitive Surgical, Sunnyvale, Calfornia), which allow the surgeon complete control over the robot remotely; and (3) shared-control models with varying levels of simultaneous control between the robot and the surgeon. Most contemporary robotic systems utilized in spine surgery today are shared-control systems which allow the planning of stereotactic trajectories via preoperative or intraoperative imaging, which is then produced by the robotic arm followed by placement of the instrumentation by the surgeon.
In terms of volume, there remains significant potential in spine surgery for the utilization of robotic and navigation platforms that could justify the initial acquisition costs. From 2004 to 2015, the volume of elective lumbar spinal fusions increased 62.3% from 122,679 cases (60.4 per 100,000) in 2004 to 199, 140 (79.8 per 100,000) in 2015. The first commercially available robotic platform for spine surgery, the Mazor SpineAssist system (Mazor Robotics, Caesarea, Israel), gained FDA approval in 2004. However, it did not see significant utilization until 2011 with a small number of installations; fewer than 100 were installed by 2015. The number of procedures per system has been increasing steadily since then, implying greater penetration of spine robots in the operating room (OR), and, by 2015, over 3000 robotic spine procedures were performed annually in the United States. A similar trend has been noted for the da Vinci Surgical System in prostate and OB/GYN surgeries, with over 700,000 procedures performed in 2015. Apart from being the first robotic system to be utilized in surgical care, the da Vinci robot also served to popularize robotic surgery both among medical professionals and the general public. Following the integration of these technologies into their daily surgical practice, if hospitals responsibly market themselves as state-of-the-art spine surgery centers, they could potentially leverage the high surgical volume to offset the initial acquisition and maintenance costs while simultaneously improving patient safety outcomes and satisfaction. Improvement of patient understanding of the surgical procedure is also essential to allow increased utilization. The patients should be informed that the surgeon would still perform the critical aspect of the procedure with the robot serving as a “guidance” mechanism for improved trajectory planning and kinetic feedback.
While most of the current literature on robotics and navigation focuses on the placement of pedicle screws, there are multiple other applications that could provide a justification for increased utilization. Several studies have demonstrated the feasibility of image guidance systems in the surgical workflow for minimally invasive lateral lumbar interbody fusion (MIS-LLIF). Both navigation and robotics have also been investigated for the placement of S2-alar-iliac screws. The applications for these systems have also expanded to the resection of both primary and metastatic spinal column and intradural tumors, along with cases of spinal deformity. In addition to the surgical volume, such an expansion of scope for a wide range of clinical indications also provides an inherent basis for cost-effectiveness.
Cost-Effectiveness of Image-Guided Navigation Systems
There have been several studies that have demonstrated the costs and benefits associated with use of image guidance in the OR Mainly, the currently available literature has been focused on the cost savings associated with the use of image guidance for the placement of pedicle screws. The proposed mechanisms of cost-effectiveness include a reduction in the rate of revision surgery due to screw malposition and a reduction in OR facility costs due to the shorter operating time as a consequence of the shorter time required for each pedicle screw placement. In an analysis of 100 patients who prospectively underwent thoracolumbar instrumentation with 3D fluoroscopic image guidance compared to a historical cohort of 100 patients who underwent instrumentation without image guidance, Watkins et al. demonstrated cost savings of $71,000 per 100 cases due to the reduction in rate of revision surgery from 3% to 0% and shortened time for screw placement (OR costs: $93/min). The total cost of the navigation system itself was reported as $475,000. Each revision surgery cost the hospital as much as $23,000 for Medicare and ~$40,000 for private patients. Therefore, image guidance was deemed cost-effective in the long term with high surgical volume. Further, Costa et al. demonstrated that image guidance based on intraoperative CT (O-Arm) may be more cost-effective than is preoperative CT . The authors found a shorter mean time for each pedicle screw placement using intraoperative CT-based computer guidance (16 min vs. 28 min) with fewer intraoperative radiographs needed. The mean cost of the procedure with intraoperative O-arm CT was also found to be lower than preoperative CT (€6482 vs. €6738). Hodges et al. projected that the use of intraoperative O-arm guidance could lead to total cost savings of $40,595,000 nationally, assuming a 1% rate of return to the OR due to breached pedicle screws with conventional C-arm fluoroscopy, compared to none in the O-arm group that was observed in their analysis from a single institution.
Cost-Effectiveness of Robotics in Surgery
Evidence From Other Surgical Specialties
While the literature specific to spine surgery has been sparse, there has been a mounting amount of evidence evaluating the cost-effectiveness associated with the use of robotic systems in other surgical specialties such as general surgery, OB/GYN, and urology. In a head-to-head comparison between robotic and open adrenalectomy, Probst et al. demonstrated that the total perioperative costs for the entire hospital admission were lower with robotic surgery (€7334 vs. €8625), although the immediate procedure costs were higher by €2288. Similar results obtained from evaluations of robotic-assisted laparoscopic surgery (RALS) for prostatectomy and hysterectomy have demonstrated a higher mean cost associated with these procedures. An important limitation of these studies is that these analyses are restricted to the direct OR and supply costs and the indirect costs and benefits associated with reduced loss of productivity, long-term treatment outcomes, etc. have not been evaluated. For instance, robotic-assisted prostatectomy has been shown to be associated with fewer positive surgical margins, lower use of androgen deprivation chemotherapy, and radiation within 2 years of surgery. Similarly, robotic-assisted nephrectomy, although associated with higher procedural costs, has been linked to fewer complications and reduced length of stay. It has been claimed that, on average, RALS costs 6% or $1600 more in direct costs per case performed, compared to conventional or open surgery. In the literature from these surgical specialties, most authors therefore generally conclude that RALS is less cost-effective, especially when factoring in the capital investment associated with initial acquisition (estimated at $1.75 million) and subsequent maintenance. However, the cost for these procedures has also been shown to decline with increasing surgical volume. Therefore, more robust cost-effectiveness analyses comprehensively taking into account all such relevant considerations are required to demonstrate a clear picture of RALS in the healthcare value equation. This is especially important since, if found to be less cost-effective in the absence of a clear outcome benefit, the burden of incremental cost associated with RALS may fall on the patients instead of the insurers. It has been interesting to note that, despite the high procedural costs, the utilization of RALS has increased substantially in the United States, with 2862 da Vinci systems installed by 2017. The global case volume also rose significantly from 2010 to 2017, from 136,000 to 877,000 cases, with a total market revenue of $3.1 billion for the manufacturer.
Evidence From Spine Surgery
In contrast to other surgical specialties, robotic surgery has been claimed to be cost-effective for spine surgery. However, there is a dearth of literature directly examining the cost data associated with robotic spine surgery. The only study to date has been performed by Menger et al., who analyzed cost data from 557 cases of elective spine surgery with thoracolumbar instrumentation at a single institution over a 1-year period, out of which 58 (10.4%) were minimally invasive (MIS) fusions. Although the authors did not perform any robotic cases, they attempted to analyze the projected impact on procedural distribution and costs, had robotic surgery been adopted using previously published national data. Upon review of the individual cases, it was found that the utilization of robotic assistance would allow the conversion of an additional 50 cases of open procedures (10%) to be performed as MIS; it would also be feasible to utilize robotic assistance in 337 cases (68% of open procedures). The authors used nationally available data to estimate the cost savings based on the incidence of screw malposition, operative time, reoperation rate, and length of stay. It was found that the utilization of robotic technology in the busy academic practice could have allowed annual cost savings of $608,546, with the majority of those savings attributed to the avoidance of revision surgery ($314,661) and the conversion of open procedures to MIS ($251,860), with consequent shortening of length of stay. We individually discuss below the possible reasons for the cost-effectiveness of robotic spine surgery.
Revision Rate
Reoperations are a significant cost contributor to the overall cost of spine surgery. Revision surgery within 30 days of an index operation may not be covered by medical insurance and the cost incurred may have to be borne by the provider and institution. The cost of a single revision surgery may be as high as ~$40,000. Robotic assistance has significant potential to reduce the incidence of reoperations due to the improved accuracy and precision of screw placement. In an analysis of 960 pedicle screws performed with robotic guidance, Hu et al. demonstrated an accuracy of 98.9%. In a similar fashion, Zahrawi et al. demonstrated 100% placement accuracy for percutaneous pedicle screws compared to 97% with free-hand placement. There have been multiple such studies that have demonstrated superior accuracy of pedicle screw placement with robotic assistance. It is expected that the addition of CT-based navigation to robotic guidance would further substantially improve precision and accuracy. The incidence of reoperation due to inaccurate screw placement with robotic technology is estimated at 0%. We have had similar experience with the first 50 cases in our institution, with no cases of thoracolumbar spinal instrumentation performed with robotic guidance requiring revision surgery due to screw breach. Studies have also shown that pedicle screw placement with robotic surgery avoids damage to the facet joint capsule, which has been associated with a higher likelihood of adjacent segment disease. In addition to pedicle screw placement, robotic assistance has been demonstrated to improve the accuracy of placement of interbody implants, vertebroplasty, and vertebral body biopsies substantially. In addition, multiple studies have validated its accuracy for the placement of S2-alar-iliac screws. Complications such as wound infections necessitating revision surgery have also been found to be lower in robotic-guided MIS compared to fluoroscopy-guided surgeries. A multicenter prospective study using the Mazor Robotics Renaissance Guidance system demonstrated a five-fold reduction in incidence of surgical complications compared to the fluoroscopic-guided arm.
Radiation Exposure
A reduction in radiation exposure due to minimizing the use of intraoperative fluoroscopy could contribute to long-term cost savings. Radiation exposure is usually measured as the number of seconds of fluoroscopy exposure per screw placed. In a comparison between robot-assisted open/percutaneous surgeries to conventional free-hand procedures, Kantelhardt et al. demonstrated a fluoroscopy time of 43 seconds per screw for robotic-guided open and 27 seconds per screw for robotic-guided percutaneous screw placement compared to a mean time of 77 seconds per screw for conventional surgeries. While a small number of studies have demonstrated little to no difference in radiation exposure, the vast majority of literature currently supports the hypothesis that radiation exposure as measured by fluoroscopy time is lower with robotic surgery. It must be noted, however, that the amount of radiation exposure is significantly dependent on the level of surgeon variability in intraoperative practices. The literature so far only covers those robotic systems that do not have real-time navigation as an adjunct. With the newer generation of robotic systems that combine CT-based navigation with the robotic arm, the use of intraoperative CT for planning would consequently increase the amount of radiation exposure for the surgical team, as a single O-arm spin can add up to 20 s of fluoroscopy time. Thus, more multicentric studies with a diverse group of spine surgeons with varying surgical practices and those that evaluate radiation exposure with newer generation of robotic guidance systems are needed.
Operating Room Time
The cost of the OR facility would be directly proportional to the total operative time. A shorter operative time is not only important for cost savings directly by reducing facility costs, but also for the indirect costs incurred due to the higher risk of surgical complications with the longer operative time. Several studies have demonstrated a much shorter time taken per screw placement with robotic assistance than with free-hand procedures, which leads to an overall shorter OR time. The cost of the OR facility per minute can be highly variable between institutions and can range anywhere from $15 to $100. In their analysis of a single-institutional spine practice, assuming an OR cost of $18/min, Menger et al. estimated that they could save $5713 on the 57 MIS cases performed at their institution over 1 year. Similarly, Watkins et al. estimated that their OR cost at $93/min and a reduction in OR time of 20 to 30 min could result in cost savings of $1860–$2790 per surgery.
Surgeon Ergonomics and Efficiency
Spine surgery is a physically demanding specialty as it relies upon meticulous fine motor skills. The removal of bone and placement of instrumentation not only requires physical strength on its own, but operating in the vicinity of critical neural elements simultaneously requires tremendous dexterity and precision. This becomes even more challenging while working in narrow operative corridors with MIS procedures. Given the possibility of surgeon physical and mental fatigue with long and arduous procedures, ergonomics in the OR is also an important concern. The incidence of neck pain and shoulder pain among surgeons has been shown to be as high as 40%. Therefore, spine surgery may be an ideal area where the utilization of robotics and navigation can allow ergonomically efficient procedures and prevent surgeon fatigue. This has already been shown in other surgical specialties such as cardiac surgery, OB/GYN, and urology.
Future Considerations
A variety of measures may be taken to ensure greater clinical application and improve the cost-effectiveness of robotic and navigation platforms. Future systems should allow co-registration with different imaging modalities such as MRI (in addition to CT) to allow greater visualization of soft tissue tumor borders and adjacent anatomy, including vascular structures and neural elements. This would improve the possibility of application for the resection of spinal tumors. Allowing spine robots greater capability in bony resection and handling soft tissues would further enhance clinical penetration. An improvement in cost-effectiveness would be directly dependent on further improvement in the mechanisms of cost savings already highlighted above, including a reduction in OR time and revision rate. There is little doubt that there can be a significant learning curve with the adoption of the new technologies; hence, the steps taken to bend the slope of this curve would improve OR efficiency. These steps can include comprehensive orientation and skill development sessions for surgeons before actual adoption into daily workflow. Early adoption of these technologies to the further training of residents would also allow them to be more comfortable and efficient with their use when they start independent practice. Given the variability in internal costing, it must be remembered that cost savings would be highly variable between institutions. Therefore, providers and hospitals must conduct a priori cost-effectiveness analyses before acquiring such platforms with a careful consideration of institutional surgical volume, case mix, and revision and complication rates from their own experience. Combining this with the already published data on the outcomes following robotic surgery can help estimate the projected cost savings before acquiring a new system. There is little doubt that cost-effectiveness can only be demonstrated in the long term as the upfront acquisition costs remain high. The “number needed to treat” and consequent time duration required for cost-effectiveness goals to be met would be highly dependent on institutional practices and case volume.
Conclusion
Despite the prohibitive upfront cost of acquiring robotic and navigation platforms for spine surgery, there remains potential for cost savings in the long term with a reduction in rate of revision surgery, operative time, radiation exposure, and improvement in surgical ergonomics. Given the lack of evidence on this important aspect, further studies that directly examine internal costing data are required to inform current spine surgical practices about cost-effectiveness and therefore estimate the value of incorporating these new technologies for patient care.