The use of robotics in spinal surgery and the medicolegal implications

Advances in technology have led to the development of computer-assisted techniques that provide visual, navigational and mechanical assistance in many fields of surgery. The use of robotics in spinal surgery is relatively new, despite it being used for many years in other specialities (such as urology). The first commercially available positioning device for the placement of the lumbar pedicle screws used in spinal fusion received approval in the USA in 2004. Since then, the use of computer-assisted surgery has expanded to include spinal decompression, implant placement and minimally invasive techniques.  

There are currently two main systems available. Optical navigation-based systems typically have a floor-mounted mechanical arm, which is used to track reference markers attached to the patient determined by intraoperative 3-dimensional (3-D) imaging. In contrast, automated anatomy recognition-based robotic guidance systems are bed- or patient-mounted and connect directly to the patient’s bony anatomy. They rely on preoperative planning using high-resolution 3-D imaging. Automated anatomy recognition software recognises individual vertebrae and uses two fluoroscopy images to merge the segmented spine to the 3-D images and the patient’s location relative to the robotic arm, with each vertebra being individually registered. Bed mounted systems are more commonly used and thus there is more information about their safety and effectiveness.

Robotics confer several advantages when used in spinal surgery. First, the accuracy of screw placement is enhanced compared to those inserted freehand, with some studies reporting a misplacement rate approaching zero. Second, exposure to fluoroscopy is significantly reduced, benefitting not only the patient but also the surgeon and other operating theatre personnel. This reduction may be as much as 80% per screw, with total intraoperative exposure to radiation per case in patients operated on with a bed-mounted system being less than half that of fluoroscopy-guided surgery. The Mazor system is an example of this and requires patients to undergo pre-operative planning consisting of a computed tomography (CT) scan, which is itself associated with radiation exposure. Therefore, the magnitude of this potential benefit remains unclear. Finally, the risk of unresolved complications is significantly lower than in comparison surgical groups and this advantage is maintained for up to 1 year post-surgery. Similarly, revision surgery, for conditions such as pseudoarthrosis, back pain requiring instrumentation removal, and radiculopathy requiring either decompression or screw removal, is required less frequently in patients operated on with a Mazor system.  

It is unlikely that the lower rates of complications and revision surgery associated with the use of robotic systems can be solely attributed to the accuracy of this type of instrumentation. Instead, the detailed 3-D planning required increases the surgeon’s familiarity with each patient’s individual anatomy, thus reducing the occurrence of surprise findings during surgery. This preoperative planning also allows the implant size and trajectory to be optimised to each patient’s anatomy. Patient-specific simulation for ideal screw trajectory can then be reproduced accurately and reliably by the robot during the actual procedure.  

The precise placement of implanted screws prevents neurologic and vascular damage, while at the same time providing proper fixation and stability to support bone fusion. However, this is difficult to achieve in minimally invasive surgery as the surgeons must rely on indirect visualisation of the patient’s anatomy, as provided by imaging systems. Screws placed using robotic systems have been shown to be associated with fewer proximal facet joint violations and better orientation. This may reduce the incidence of strain being placed on pedicle screws during rod insertion, an issue that can impact bone purchase, soft tissue tension and problematic wounds.  

There is emerging evidence that use of a Mazor system can also alleviate the stress on adjacent vertebrae compared to pedicle screws inserted by the freehand technique. This may result from optimisation of the relationship between patient anatomy and implant placement, as well as the ‘single pass’ drilling of the pilot hole offering better bone purchase.  

As with all technology, robotics are not infallible. Surgical procedures may have to be aborted due to failure of the registration software or a lack of adequate fluoroscopic images. Soft tissue pressure can lead to mispositioning of the guiding arm and inaccurate screw placement. Additionally, it may not be surgically possible to achieve the necessary angles of placement indicated by the registration software. Screw misplacement appears to be lower in patients in whom the preoperative CT scan is performed in the prone position compared to those taken in the more usual supine position. Some studies have also described difficulty in keeping the drill guide in position on the slope of the facet, leading to lateral and inferior deviations. In this situation, accuracy of screw placement can be ensured by securely docking the drill guide to prevent movement while drilling the pilot hole. Finally, much of the available evidence on the efficacy and safety of robotic systems comes from studies of single-level fusion and it is unclear how applicable this is to more complex scenarios.  

Although still in its infancy, the use of robotics in spinal surgery has the potential to mitigate some of the current drawbacks associated with these procedures by improving the accuracy of screw placement and removing the element of human error associated with manual techniques. Exposure of patients and hospital staff to radiation may also be reduced. The reduction in subsequent complications and reduced need for revision surgery should have a positive impact on patient outcomes. There is still a long way to go with this technology, and rather than a robot performing surgery of its own accord, its main function is to provide assistance to the surgeon to improve the accuracy of metalwork placement.