Robotic Spine Surgery for Revision and Complex Cases

Robotic spine surgery for revision cases uses a CT scan to plan the exact path of every pedicle screw before the operation, and a robotic arm to line up the drill along that path. A fixed plan helps most when earlier surgery has left scar and old implants that hide the usual landmarks.

The technique is built for bone work: fusion, deformity, revision and upper cervical fixation. Dr. Puneet Girdhar, Chairman - Max Institute of Robotics and Minimal Invasive Spine Surgery - Ortho Spine, MIS & Robotic Spine Surgery, BLK-Max Super Speciality Hospital, New Delhi, India, introduced it to his department about four and a half to five years ago and has since used it in more than 500 patients. He reports that every screw so far has been placed centrally in the pedicle, in thousands of screws.

How a robotic spine operation runs

Robotics in spine surgery works only on the bone model, so the process starts with a CT scan. Dr. Girdhar's theatre has an intraoperative scanner, so the patient is scanned in theatre with a reference frame placed on the area to be operated.

The 3D scan appears on the robot's planning software on four screens: a three-dimensional view and the coronal, sagittal and axial sections. The surgeon drags and drops each screw onto the images and adjusts the entry point and trajectory in every plane. This takes a couple of minutes for the whole construct. The surgeon then commands the robot to line up for each screw in turn, and the drilling and screw insertion follow that line.

During insertion, the screen shows the planned trajectory in blue and the executed trajectory in yellow. Dr. Girdhar says the two almost always overlap, which he describes as a bullseye. The 3D image can be rotated virtually, which lets him confirm that the screw stays in bone throughout, without entering the foramen above or below or the spinal canal, and that the entry point does not collide with the iliac crest (the upper rim of the pelvis).

Three practical consequences follow:

• Radiation: the only radiation is the pre-operative or intraoperative 3D CT. After that, screws are placed on navigation feedback, and no repeated check with the C-arm (the mobile X-ray unit) is required while robotic screws are inserted.
• Speed: Dr. Girdhar says 20 screws can be placed in 10 minutes.
• Reach: the technique is not restricted to any number of levels.

What 100% accuracy means

Dr. Girdhar reports a 100% accuracy rate in his series of more than 500 robotic patients. He refers to the Gertzbein-Robbins criteria, under which grade A means that the screw lies entirely inside the pedicle. Placing screws at this standard matters because a screw that breaches the pedicle wall can injure a nerve root, the dura or, in the neck and chest, vessels.

The figure is his own report from his own centre. It is higher than the grade A rates usually reported in published robotic series, so it should be read as the result of one unit, not as a guarantee attached to the technology.

Why screw placement is the step worth automating

A pedicle screw is driven through a narrow bony channel with the spinal canal on one side and the nerve root exit on the other. Placed freehand, it depends on the surgeon's anatomical knowledge, feel and repeated X-ray checks. Published series consistently report that a meaningful minority of freehand screws breach the pedicle wall, and that navigated and robotic techniques lower the rate. The consequence of a breach ranges from none to nerve root irritation or injury, and screw position is routinely checked on imaging after surgery.

Navigation and robotics differ in what they add. Navigation shows the surgeon where the instrument is relative to the scan. A robot goes one step further by holding the guide in the planned line while the drill advances. In Dr. Girdhar's workflow the two operate together: the scan is displayed through the navigation software, the plan is drawn on it, and the robot executes it.

Robotic spine surgery for revision cases: where a plan helps most

The unit performs 20 to 30 revision spinal operations a month, and Dr. Girdhar says robotics gives an extra edge in them, because the trajectory can be pre-planned. In a patient who has been operated on before, the anatomy is altered by scar and by old implants, and the landmarks a surgeon depends on can be hard to find. With the plan made in advance, the trajectory is already chosen, so the needle does not have to be located by feel in scarred tissue where it could slip. In a previously operated patient, the plan also shows where the iliac crest lies, so that the screw trajectory is not obstructed.

The unit also uses expandable cages for reduction and for restoring lumbar lordosis (the normal inward curve of the lower back), and it performs its reductions of spondylolisthesis, from grade 1 to grade 4, under image guidance. Dr. Girdhar treats at least 300 patients with spondylolisthesis every year.

Upper cervical fixation: the C1-C2 transarticular screw

Another application he showed is in the upper neck. A C1-C2 transarticular screw crosses the joint between the first and second cervical vertebrae, and the vertebral artery runs close to its path. With robotic planning, the surgeon can place the screw so that it avoids the artery and stays out of the spinal canal. The planned trajectory is fixed in advance and a power drill is used along it. Dr. Girdhar has performed the procedure with the robot.

The learning curve and the team

Dr. Girdhar's view is that the robot flattens the learning curve. Instead of relying on eyeballing and hand-eye coordination, a fellow can follow the same objective path the plan defines. He is also clear that every technology has its own drawbacks, and that the learning curve for robotics exists, though it is very short compared with learning eyeballing skills and hand-eye coordination.

What a robot does not do

A robot is a tool for placing implants in bone, and its scope ends there. It does not decompress a nerve, which is soft-tissue work done under a microscope or an endoscope, and it does not choose the operation: that decision depends on diagnosing the problem correctly, as discussed in Indications for Spine Surgery: A Complete Guide.

Dr. Girdhar is explicit about a further limit. Patients sometimes ask whether robotic surgery can improve a complete spinal cord injury of long standing, with no sensation, no power in the legs and no bladder or bowel control. His answer is no: robotic surgery is meant to place implants accurately and to prevent neurological injury, and it does not restore the nerves.

Connected reading

For the unit's wider approach and its safety record, read Endoscopic and Robotic Spine Surgery: A Complete Guide. The endoscopic side of the work is in Biportal Endoscopic Spine Surgery (UBE): Discs to TLIF, and bone cutting for deformity and revision is in Ultrasonic Osteotome and Spinal Osteotomy Types Explained.

Dr. Girdhar's August 2025 Jivo Masterclass explained how navigation and robotics improve accuracy and safety in How Robotic Spine Surgery Works: Planning, Accuracy, Safety; this article concentrates on his results in revision and complex cases. The Live Discussion records the questions from this session, and his profile is on the doctor page. For budgeting, see the guides to spinal decompression and fixation and scoliosis and spinal deformity correction.