When Scoliosis Surgery Doesn’t Go As Planned — Revision Surgery in New York/New Jersey

Dr. Zeeshan Sardar, MD, MSc, F.R.C.S.C
Co-Chief of Spinal Deformity Surgery • Director, Quality & Patient Safety • Medical Director, Spine Unit
Och Spine Hospital at NewYork-Presbyterian / Columbia University • New York, NY

Scoliosis surgery is successful for the large majority of patients. But a meaningful minority — published estimates range from 10 to 20 percent — experience a complication or outcome that is not what was expected: persistent pain, visible imbalance, a curve that has continued to progress, hardware that has broken, or a fusion that never fully solidified. These patients frequently find themselves in a difficult position. They return to their original surgeon and are told that the X-rays look acceptable. They are advised that ongoing discomfort is normal, that they need to give it more time, or that nothing further can be done. Some accept this and learn to live with it. Others, correctly, suspect that something is not right — and that a second evaluation is appropriate.

This page is written for those patients. It explains the specific ways scoliosis surgery can fail to deliver its intended outcome, what symptoms and imaging findings should prompt evaluation, and what revision surgery — when appropriate — involves.

THE EXPERIENCE MANY PATIENTS SHARE

Patients who come to see Dr. Sardar after prior scoliosis surgery often describe a recognizable experience: they had surgery, were told the procedure went well, were discharged with instructions to give it time, and then — months or years later — continued to be symptomatic in ways that don’t match the reassurance they received. Common variations include:

  • Persistent back pain or leg pain well beyond the expected recovery period, with the original surgeon attributing it to normal post-surgical changes
  • Visible imbalance — a shoulder that is still uneven, a trunk that leans to one side, a posture that never fully corrected — that the patient can see but is told is acceptable
  • A curve that appears to be progressing above or below the fusion on serial X-rays, with uncertainty about whether this is expected or concerning
  • A broken rod or loose hardware identified on imaging, with conflicting advice about whether or not it needs to be addressed
  • Neurological symptoms — arm or leg weakness, numbness — that developed after surgery or have worsened since

None of these situations is automatically a surgical emergency. But all of them deserve a thorough evaluation.

THE SPECIFIC WAYS SCOLIOSIS SURGERY CAN FAIL

1. Pseudarthrosis — Failed Fusion

Pseudarthrosis is the failure of the bone graft to consolidate into solid fusion. The hardware looks intact on plain X-rays, the curve appears corrected, and the patient was discharged with a technically successful surgery note — but the bone never fused. Instead, a fibrous non-union persists at one or more levels within the fusion mass, creating a zone of micromotion that causes pain and, over time, hardware stress fracture.

Pseudarthrosis is the most common indication for revision after scoliosis surgery in published series. It is often underdiagnosed because standard X-rays can appear reassuring even when the fusion has not solidified. CT scan with fine cuts through the fusion mass is the gold standard for diagnosing pseudarthrosis. A broken rod at a specific level — particularly a clean transverse fracture through the rod at a single level — is often the imaging marker that reveals an underlying pseudarthrosis at that segment.

Published revision series report bony union rates of approximately 98 percent with appropriate revision surgery for pseudarthrosis — meaning the underlying problem is eminently fixable for patients who receive the right diagnosis and treatment.

2. Hardware Failure

Pedicle screws, rods, hooks, and cross-links are the mechanical infrastructure of a spinal fusion. They are designed to hold alignment while bone heals and then to function alongside the fusion mass long-term. They can fail in several ways:

  • Rod fracture — the most common hardware failure, almost always occurring at a level of underlying pseudarthrosis; the rod fails because it is absorbing cyclic loading that solid bone should be absorbing instead
  • Screw pullout or loosening — particularly in patients with osteoporosis or in areas of poor bone quality; loose screws may cause localized pain and represent a potential threat to adjacent neural structures if they migrate
  • Hook migration — a late complication of older-generation hook-based instrumentation, where a hook at the top or bottom of the construct migrates and may compress neural structures; particularly relevant for patients who had surgery in the 1990s and early 2000s
  • Implant prominence — hardware that becomes symptomatic as soft tissue thins over it with age, requiring removal even in the absence of fusion failure

Not all hardware failure requires immediate revision. A broken rod in an otherwise asymptomatic patient with a solid fusion may be monitored rather than revised. The decision depends on the specific implant, its location, the patient’s symptoms, and whether the underlying fusion is solid.

3. Coronal Imbalance

Coronal imbalance refers to the trunk sitting visibly off-center over the pelvis when viewed from the front — a trunk shift, a persistent lean, or significant residual shoulder asymmetry. Some degree of residual asymmetry is expected and acceptable; surgical goals for scoliosis are balance and stability, not radiographic perfection. But significant coronal imbalance — a trunk shifted more than 3 centimeters to one side — should be evaluated.

Patients with significant postoperative coronal imbalance often describe difficulty walking in a straight line, progressive hip or knee pain from asymmetric loading, and cosmetic concerns that significantly affect their quality of life. Evaluation requires careful standing X-rays measuring the coronal plumb line and trunk shift, comparison with preoperative imaging, and assessment of whether the imbalance is structural or compensatory.

4. Sagittal Imbalance and Flatback Deformity

Sagittal imbalance — specifically the loss of lumbar lordosis that produces a progressive forward lean — can develop after scoliosis surgery when the fusion fails to restore or maintain appropriate lumbar alignment. This is the defining complication of Harrington rod surgery from prior decades, but it can also occur after modern instrumented fusion when lordosis restoration is inadequate. The clinical picture — progressive difficulty standing upright, back and leg fatigue, knee-bent compensatory posture — is discussed in detail on the Flatback Deformity Surgery page.

5. Curve Progression — Adding-On and Junctional Failure

Scoliosis correction fuses a defined segment of the spine and leaves the levels above and below mobile. After surgery, several patterns of curve-related failure can emerge:

  • Adding-on below the fusion — progressive curve development below the lower instrumented vertebra, typically from selection of a fusion endpoint that did not adequately include the structural curve. The unfused lumbar spine progresses as the residual structural curve is no longer controlled. Adding-on can occur months to years after surgery and is often the indication for early revision by extension of the fusion construct.
  • Proximal junctional kyphosis (PJK) — new kyphotic angulation at the levels immediately above the top of the fusion; discussed in detail on the Scheuermann’s Kyphosis page in context of kyphosis surgery, but it occurs after scoliosis fusion as well. Most PJK is mild and radiographic; symptomatic PJK with neurological involvement or severe progressive deformity warrants revision consideration.
  • Curve progression in unfused segments — compensatory curves that were mobile at the time of surgery may develop into structural curves over time, particularly in adult patients whose degeneration continues. Monitoring with periodic standing X-rays is appropriate for all post-scoliosis-surgery patients to identify this pattern early.

6. Late Infection

A subset of patients with persistent pain after scoliosis surgery — often months to years after a procedure with no acute postoperative infection — harbor a low-grade, indolent infection that standard inflammatory markers and imaging may not detect. Cutibacterium acnes (formerly Propionibacterium acnes) is a slow-growing skin organism that can contaminate spinal implants at the time of surgery without producing the acute wound findings typical of early infection. It causes chronic, low-grade inflammation around implants that manifests as persistent pain, without fever or obvious wound changes.

Late infection is rare but important to consider in patients with otherwise unexplained persistent pain after scoliosis surgery, particularly when multiple other causes have been excluded. Diagnosis requires detailed sampling at the time of revision surgery, with extended culture periods specifically for slow-growing organisms.

ADOLESCENT VS. ADULT SCOLIOSIS REVISION — DIFFERENT PATIENTS, DIFFERENT CONSIDERATIONS

Adolescent Scoliosis Revision (AIS)

Many of the patients presenting for scoliosis revision evaluation are young adults in their 20s, 30s, or 40s who had AIS surgery as teenagers. The failure patterns most common in this group are pseudarthrosis, hardware failure, coronal imbalance, and adding-on. These patients are generally in good overall health and are otherwise excellent surgical candidates for revision — but the presence of prior instrumentation throughout the fusion segment, altered anatomy from the index surgery, and occasionally dense scar tissue from the prior approach all make revision more demanding than a primary deformity correction would have been.

The emotional dimension for AIS revision patients is also specific: many of these patients have lived with the consequences of a less-than-optimal outcome since adolescence, have built a picture of their situation as permanent, and need clear and honest communication about what revision can and cannot realistically achieve.

Adult Scoliosis Revision

Adult scoliosis revision is generally more complex than AIS revision for several reasons: the initial surgery was almost certainly longer and more extensive, involving more vertebral levels and often more osteotomy; the patient’s bone quality has typically declined since surgery; adjacent segment degeneration and adjacent segment disease are more prevalent in adult deformity patients; and the range of failure patterns is wider. Adult scoliosis revision patients are also more likely to have systemic conditions — osteoporosis, diabetes, cardiovascular disease — that affect surgical planning and risk stratification.

Revision of adult deformity surgery requires the same preoperative evaluation framework as primary adult deformity surgery — full-length standing X-rays, CT, MRI, bone density assessment — plus detailed review of all prior operative records to understand what was done at each prior procedure and what anatomy is available for revision instrumentation.

WHAT A REVISION EVALUATION INVOLVES

A thorough revision evaluation for a patient with prior scoliosis surgery requires:

  • Prior operative reports — knowing what was done at the index surgery (levels fused, hardware used, osteotomies performed, blood loss, complications noted) is important context for interpreting the current imaging
  • Serial postoperative X-rays — the trajectory of alignment over time is important for assessing whether curves are stable or progressing
  • Full-length standing X-rays (36-inch cassette) in the current state — measuring coronal and sagittal balance, assessing the fusion construct, and evaluating unfused levels above and below
  • CT scan — to assess fusion mass integrity (diagnosing pseudarthrosis), hardware condition, and the bony anatomy available for revision instrumentation
  • MRI — to assess neural structures, disc levels above and below the fusion, cord signal, and any component of nerve compression contributing to symptoms
  • Bone density assessment — critical for adult and older patients; affects implant selection and the decision about preoperative optimization with bone-building medications

Bringing this complete set of records and imaging to a consultation — rather than only the most recent imaging — allows the most thorough evaluation at a single visit. Telemedicine consultations are available for patients who want to evaluate options before traveling to New York.

WHY REVISION SCOLIOSIS SURGERY IS DEMANDING

Revision scoliosis surgery is categorically more difficult than primary surgery. Several factors compound in the revision setting:

  • Prior instrumentation must be removed or worked around. Removing pedicle screws from a solidly fused segment is technically demanding; in some cases it is safer to leave portions of prior hardware in place and augment rather than attempt complete removal.
  • Scar tissue from the prior surgery. Each prior posterior approach leaves epidural and paraspinal scarring that must be meticulously managed to avoid neural injury during the revision dissection.
  • Bone quality is typically diminished at revision. Previously instrumented pedicles may have enlarged screw tracts, reduced cortical integrity, or bone loss around prior hardware sites — all of which must be accounted for in the revision fixation strategy.
  • Blood loss is higher. Revision surgery through scarred tissue is associated with greater blood loss than primary surgery; cell salvage and proactive transfusion planning are essential.
  • The volume of revision experience matters. A surgeon who operates primarily in primary deformity and encounters revision cases infrequently is not working from the same foundation as a surgeon for whom revision represents a substantial, routine portion of the practice.

At NewYork-Presbyterian / Columbia University, a significant proportion of Dr. Sardar’s operative cases are revisions — including re-revisions referred from other surgeons. The full institutional infrastructure — robotic navigation, continuous intraoperative neuromonitoring, intraoperative CT, cell salvage, and critical care — is available as standard for all revision cases.

FREQUENTLY ASKED QUESTIONS

How do I know if my scoliosis surgery failed?

There is no single definition of scoliosis surgery failure — it exists on a spectrum. Persistent pain that has not improved over 12 to 18 months, documented curve progression on serial imaging, visible imbalance, neurological symptoms, or broken hardware are all findings that warrant formal evaluation rather than reassurance. The absence of acute emergency symptoms does not mean a concerning finding should be watched indefinitely without expert review.

My X-rays were called normal. Can the surgery still have failed?

Yes. Plain X-rays are not sensitive for pseudarthrosis. A CT scan with fine-cut sequences through the fusion mass is required to diagnose or exclude pseudarthrosis. Patients with persistent pain after scoliosis surgery who have been told their X-rays look fine deserve a CT scan before that reassurance is accepted as definitive.

Should I go back to my original surgeon or see someone new?

This is a personal decision. Returning to the original surgeon has advantages — they know your anatomy, have your records, and may be best positioned to address specific technical issues. But if you have returned to your original surgeon and your concerns have not been addressed to your satisfaction, or if the original surgeon is not available or willing to engage further, a second opinion from an independent revision specialist is entirely appropriate. Dr. Sardar regularly evaluates patients who have had prior scoliosis surgery elsewhere, with no prior involvement in the original case.

How long after scoliosis surgery can problems develop?

The timeline varies by failure type. Pseudarthrosis and hardware failure can manifest from months to years after surgery. Adding-on typically becomes evident in the first few years of follow-up. Adjacent segment disease develops over a decade or more. Late infection can manifest years after a technically uncomplicated index procedure. The common thread is that scoliosis surgery requires lifelong periodic monitoring — not just through the immediate recovery period but with standing X-rays at appropriate intervals over the long term.

Is revision surgery riskier than my original procedure?

Yes, in general terms. The presence of prior instrumentation, scar tissue, diminished bone quality, and more complex planning all make revision surgery more demanding than primary surgery. However, the comparison that matters for decision-making is not between revision and no surgery, but between revision and continued deterioration from the unaddressed failure. For patients with pseudarthrosis, progressive imbalance, or neurological compromise, the risk of not revising is typically greater than the risk of proceeding with expert revision at a high-volume center.

Can I be evaluated via telemedicine first?

Yes. Many patients traveling from other states for a revision evaluation start with a telemedicine consultation, during which imaging is reviewed and the clinical picture is assessed. This allows a meaningful first evaluation before any travel commitment. Telemedicine consultations are available for patients in NY, NJ, CT, FL, PA, MO, CA, and TX.

RELATED CONDITIONS & PROCEDURES

About Dr. Zeeshan Sardar
Dr. Sardar is Co-Chief of Spinal Deformity Surgery at NewYork-Presbyterian / Columbia University. He has no prior involvement in the original scoliosis surgery of patients who come to him for evaluation — his assessment is entirely independent. A significant proportion of his practice consists of revision cases referred from other surgeons and institutions, including re-revisions. To schedule a consultation or request a second opinion, call 212-932-5187 or visit the contact page. Telemedicine available in NY, NJ, CT, FL, PA, MO, CA, and TX.

This page is for educational purposes only and does not constitute individualized medical advice. Please consult a qualified spine specialist to discuss your specific situation.

Reviewed by Dr. Zeeshan Sardar, MD, MSc, F.R.C.S.C — Last reviewed: August 2026.