Development of a Novel Cervical Deformity Surgical Invasiveness Index

Peter G. Passias, MD; Samantha R. Horn, BA; Alexandra Soroceanu, MD; Cheongeun Oh, PD; Tamir Ailon, MD; Brian J. Neuman, MD; Virginie Lafage, PhD; Renaud Lafage, MS; Justin S. Smith, MD; Breton Line, BS; Cole A. Bortz, BA; Frank A. Segreto, BS; Avery Brown, BS; Haddy Alas, BS; Katherine E. Pierce, BS; Robert K. Eastlack, MD; Daniel M. Sciubba, MD; Themistocles S. Protopsaltis, MD; Eric O. Klineberg, MD; Douglas C. Burton, MD; Robert A. Hart, MD; Frank J. Schwab, MD; Shay Bess, MD; Christopher I. Shaffrey, MD; Christopher P. Ames, MD


Spine. 2020;45(2):116-123. 

In This Article

Abstract and Introduction


Study Design: Retrospective review.

Objective: The aim of this study was to develop a novel surgical invasiveness index for cervical deformity (CD) surgery that incorporates CD-specific parameters.

Summary of Background Data: There has been a surgical invasiveness index for general spine surgery and adult spinal deformity, but a CD index has not been developed.

Methods: CD was defined as at least one of the following: C2-C7 Cobb >10°, cervical lordosis (CL) >10°, cervical sagittal vertical axis (cSVA) >4 cm, chin brow vertical angle >25°. Consensus from experienced spine and neurosurgeons selected weightings for each variable that went into the invasiveness index. Binary logistic regression predicted high operative time (>338 minutes), estimated blood loss (EBL) (>600 mL), or length of stay (LOS) >5 days) based on the median values of operative time, EBL, and LOS. Multivariable regression modeling was utilized to construct a final model incorporating the strongest combination of factors to predict operative time, LOS, and EBL.

Results: Eighty-five CD patients were included (61 years, 66% females). The variables in the newly developed CD invasiveness index with their corresponding weightings were: history of previous cervical surgery (3), anterior cervical discectomy and fusion (2/level), corpectomy (4/level), levels fused (1/level), implants (1/level), posterior decompression (2/level), Smith-Peterson osteotomy (2/level), three-column osteotomy (8/level), fusion to upper cervical spine (2), absolute change in T1 slope minus cervical lordosis, cSVA, T4-T12 thoracic kyphosis (TK), and sagittal vertical axis (SVA) from baseline to 1-year. The newly developed CD-specific invasiveness index strongly predicted long LOS (R 2 = 0.310, P < 0.001), high EBL (R 2 = 0.170, P = 0.011), and extended operative time (R 2 = 0.207, P = 0.031). A second analysis used multivariable regression modeling to determine which combination of factors in the newly developed index were the strongest determinants of operative time, LOS, and EBL. The final predictive model included: number of corpectomies, levels fused, decompression, combined approach, and absolute changes in SVA, cSVA, and TK. This model predicted EBL (R 2 = 0.26), operative time (R 2 = 0.12), and LOS (R 2 = 0.13).

Conclusion: Extended LOS, operative time, and high blood loss were strongly predicted by the newly developed CD invasiveness index, incorporating surgical factors and radiographic parameters clinically relevant for patients undergoing CD corrective surgery.

Level of Evidence: 4


Cervical deformity (CD) includes a wide range of disorders and etiologies that include congenital conditions, ankylosing spondylitis, trauma, and others. As CD progresses, it can lead to severe disability and decreased neurologic functioning.[1,2] Unlike for thoracolumbar deformities, CD literature is lacking in a strong consensus on characterization and classification of CD.[3] CD surgical correction is challenging and demands preoperative planning that takes into account patient baseline functional status, sagittal alignment, comorbidities, and frailty which can all be used for preoperative risk assessment.[4] The wide variability in treatment techniques and surgical planning for CD in part contributes to the large range of outcomes and complications.[5] CD patients are a distinct entity from patients with thoracolumbar adult spinal deformity.[4] CD patients are frailer than thoracolumbar deformity patients, increasing the risk associated with operating on this cohort. Additionally, CD patients comprise more heterogeneous pathologies, and often require more complex surgeries to achieve adequate correction.[6] Complication types and rates are also different between CD and isolated thoracolumbar deformity patients, with unique complications such as C5 palsy and dysphagia.[7]

Previous studies have developed surgical invasiveness indices for both general surgery and adult spinal deformity, identifying and weighting procedure-related and alignment parameters that contribute to the overall invasiveness of a spine procedure. Mirza et al[8] developed a surgical invasiveness index to assess the invasiveness of general spine surgeries, not limited by indications for surgery. This index assigned a point value to each component of a general spine procedure, including one point per vertebra requiring decompression, one point per vertebra that has graft material attached to or replacing it, one point per vertebra with instrumentation for both the anterior and posterior approach. The component scores are then summed to generate the total surgical invasiveness score proposed by Mirza et al. This index was then validated using other measures to assess surgical invasiveness, including operative time and estimated blood loss (EBL). Neuman et al[9] then developed a surgical invasiveness index specific for an adult spinal deformity (ASD) population, given that deformity-specific factors are not taken into account in the Mirza index but contribute to the invasiveness of the procedure. Neuman et al found that their developed index that included both surgical and radiographic components specific to an ASD population predicted the invasiveness of ASD surgery better than the Mirza surgical invasiveness index.[8,9]

No studies to date have developed an invasiveness index specific to a CD population. The Mirza surgical invasiveness index is not specific to a deformity population and the Neuman ASD invasiveness index is specific to a thoracolumbar ASD population, which has some overlap with the CD patient population but is its own entity. Therefore, we aimed to develop an invasiveness index with CD-specific factors that could assess the invasiveness of CD surgeries with the goal of generating a tool that can aid in risk assessments and outcomes in a CD cohort.