Predictive analytic algorithms are designed to identify patterns in the data that allow for accurate predictions without the need for a hypothesis. Therefore, predictive modeling can provide detailed and patient-specific information that can be readily applied when discussing the risks of surgery with a patient. There are few studies using predictive modeling techniques in the adult spine surgery literature. These types of studies represent the beginning of the use of predictive analytics in spine surgery outcomes. We will discuss the advancements in the field of spine surgery with respect to predictive analytics, the controversies surrounding the technique, and the future directions.

Adult spinal deformity (ASD) surgery is still associated with high surgical risks. Machine learning algorithms applied to multicenter databases have been created to predict outcomes and complications, optimize patient selection, and improve overall results. However, the multiple data points currently used to create these models allow for 70% of accuracy in prediction. We need to find new variables that can capture the spectrum of probability that is escaping from our control. These proposed variables are based on patients biological dimensions, such as frailty, sarcopenia, muscle and bone (tissue) sampling, serological assessment of cellular senescence, and circulating biomarkers that can measure epigenetics, inflammaging, and -omics. Many of these variables are proven to be modifiable and could be improved with proper nutrition, toxin avoidance, endurance exercise, and even surgery. The purpose of this manuscript is to describe the different future data points that can be implemented in ASD assessment to improve modeling prediction, allow monitoring their response to prerehabilitation programs, and improve patient counseling.

Introduction Studies document rod fracture in pedicle subtraction osteotomy (PSO) settings where disk spaces were preserved above or adjacent to the PSO. This study compares the multidirectional bending rigidity and fatigue life of PSO segments with or without interbody support. Methods Twelve specimens received bilateral T12-S1 posterior fixation and L3 PSO. Six received extreme lateral interbody fusion (XLIF) cages in addition to PSO at L2-L3 and L3-L4; six had PSO only. Flexion-extension, lateral bending, and axial rotation (AR) tests were conducted up to 7.5 Newton-meters (Nm) for groups: (1) posterior fixation, (2) L3 PSO, (3) addition of cages (six specimens). Relative motion across the osteotomy (L2-L4) and entire fixation site (T12-S1) was measured. All specimens were then fatigue tested for 35K cycles. Results Regardingmultiaxial bending, there was a significant 25.7% reduction in AR range of motion across L2-L4 following addition of cages. Regarding fatigue bending, dynamic stiffness, though not significant (p = 0.095), was 22.2% greater in the PSO + XLIF group than in the PSO-only group. Conclusions Results suggest that placement of interbody cages in PSO settings has a potential stabilizing effect, which is modestly evident in the acute setting. Inserting cages in a second-stage surgery remains a viable option and may benefit patients in terms of recovery but additional clinical studies are necessary to confirm this.

OBJECTIVE: The goal of this study was to determine if patients with mild scoliosis and age-appropriate sagittal alignment have favorable outcomes following surgical correction. METHODS: Retrospective review of a prospective, multicenter adult spinal deformity database. Inclusion criteria: operative patients age ≥18 years, and preoperative pelvic tilt, mismatch between pelvic incidence and lumbar lordosis (PI-LL), and C7 sagittal vertical axis all within established age-adjusted thresholds with minimum 2-year follow-up. Health-related quality of life (HRQoL) scores: Oswestry Disability Index (ODI), 36-item Short Form health survey (SF-36), Scoliosis Research Society-22R (SRS22R), back/leg pain Numerical Rating Scale and minimum clinically important difference (MCID)/substantial clinical benefit (SCB). Two-year and preoperative HRQoL radiographic data were compared. Patients with mild scoliosis (Mild Scoli, Max coronal Cobb 10°-30°) were compared to those with larger curves (Scoli). RESULTS: One hundred fifty-one patients included from 667 operative patients (82.8% women; average age, 56.4 ± 16.2 years). Forty-two patients (27.8%) included in Mild Scoli group. Mild Scoli group had significantly worse baseline leg pain, ODI, and physical composite scores (p < 0.02). Mean 2-year maximum coronal Cobb angle was significantly improved compared to baseline (p < 0.001). All 2-year HRQoL measures were significantly improved compared to (p < 0.001) except mental composite score, SRS activity and SRS mental for the Mild Scoli group (p > 0.05). From the mild Scoli group, 36%-74% met either MCID or SCB for the HRQoL measures. Sixty-four point three percent had minimum 1 complication, 28.6% had a major complication, 35.7% had reoperation. CONCLUSION: Mild scoliosis patients with age-appropriate sagittal alignment benefit from surgical correction, decompression, and stabilization at 2 years postoperative despite having a high complication rate.

Objective

Extension of the posterior upper-most instrumented vertebra (UIV) into the upper thoracic (UT) spine allows for greater deformity correction and reduced incidence of proximal junction kyphosis (PJK) in adult spinal deformity (ASD) patients. However, it may be associated with chronic postoperative scapular pain (POSP). The goal of this study was to assess the relationship between UT UIV and persistent POSP, describe the pain, and assess its impact on patient disability.

Methods

ASD patients who underwent multilevel posterior fusion were retrospectively identified then administered a survey regarding scapular pain and the Oswestry Disability Index (ODI), by telephone. Univariate and multivariate analysis were utilized.

Results

A total of 74 ASD patients were included in the study: 37 patients with chronic POSP and 37 without scapular pain. The mean age was 70.5 years, and 63.9% were women. There were no significant differences in clinical characteristics, including mechanical complications (PJK, pseudarthrosis, and rod fracture) or reoperation between groups. Patients with persistent POSP were more likely to have a UT than a lower thoracic UIV (p = 0.018). UT UIV was independently associated with chronic POSP on multivariate analysis (p = 0.022). ODI score was significantly higher in patients with scapular pain (p = 0.001). Chronic POSP (p = 0.001) and prior spine surgery (p = 0.037) were independently associated with ODI on multivariate analysis.

Conclusion

A UT UIV is independently associated with increased odds of chronic POSP, and this pain is associated with significant increases in patient disability. It is a significant clinical problem despite solid radiographic fusion and the absence of PJK.

Adult spinal deformity (ASD) is a complex and heterogeneous disease that can severely impact patients' lives. While it is clear that surgical correction can achieve significant improvement of spinopelvic parameters and quality of life measures in adults with spinal deformity, there remains a high risk of complication associated with surgical approaches to adult deformity. Over the past decade, utilization of surgical correction for ASD has increased dramatically as deformity correction techniques have become more refined and widely adopted. Along with this increase in surgical utilization, there has been a massive undertaking by spine surgeons to develop more robust models to predict postoperative outcomes in an effort to mitigate the relatively high complication rates. A large part of this revolution within spine surgery has been the gradual adoption of predictive analytics harnessing artificial intelligence through the use of machine learning algorithms. The development of predictive models to accurately prognosticate patient outcomes following ASD surgery represents a dramatic improvement over prior statistical models which are better suited for finding associations between variables than for their predictive utility. Machine learning models, which offer the ability to make more accurate and reproducible predictions, provide surgeons with a wide array of practical applications from augmenting clinical decision making to more wide-spread public health implications. The inclusion of these advanced computational techniques in spine practices will be paramount for improving the care of patients, by empowering both patients and surgeons to more specifically tailor clinical decisions to address individual health profiles and needs.

Study design

Retrospective cohort study.

Objective

Determine the indications, complications, and clinical outcomes in patients requiring fusions from the cervical spine to the pelvis. Several investigators have examined fusions from the thoracic spine to the sacrum, but no similar study has been performed for cervical-to-pelvis fusions.

Methods

Patients from 2003 to 2014 with an upper instrumented vertebrae (UIV) in the cervical spine (any level) and a lower instrumented vertebrae (LIV) in the sacrum or pelvis were included in the study. Those with infectious or acute trauma-related deformities were excluded. Patient demographics, medical history, diagnosis, operative procedure, and health-related quality of life measures were analyzed. Student's t test, Kruskal-Wallis test, and χ² test were used as appropriate; significance was set at P < .05 for all tests.

Results

Fifty-five patients met inclusion criteria for the study. Average follow-up was 2.8 years. Proximal junctional kyphosis was the most common indication for cervical-to-pelvis fusions (36%). The most common UIV was C2 (29%) followed by C7 (24%). There was an average 31° correction in maximum kyphosis and a 3.3 cm improvement in sagittal vertical axis. In adults, the rate of complication was 71.4%, with a major complication rate of 39.3% and reoperation rate of 53.6%. There was significant improvement in the Scoliosis Research Society (SRS-22r) score (3.0 to 3.5; P < .01).

Conclusion

Proximal junctional kyphosis is the most common indication for patients requiring fusion to the cervical spine. Adult patients incur a significant risk of major complications and reoperations. However, significant improvement in SRS-22r outcomes are noted in these patients.

Purpose

Evaluate the presence of new-onset cervical deformity (CD) in nonoperative adult spinal deformity (ASD) patients with extended follow-up, with consideration for predictors, prevalence, and impact on patient-reported outcomes.

Methods

Retrospective review of a prospective nonoperative ASD cohort. New onset CD patients at 1- (CD-1Y) and 2-year (CD-2Y) follow-up were defined as displaying baseline cervical alignment. Univariate analyses determined differences in radiographic parameters and outcome scores of CD and maintained-cervical-alignment patients. Multivariate binary logistic regression models determined new-onset CD predictors.

Results

A total of 143 patients were included (mean age 54 years, mean body mass index 25.6 kg/m², 86% female). Cervical deformity rate was 38.5% at baseline. New-onset CD incidence at 1- and 2-year follow-up was 30.0% and 41.7%, respectively. Global sagittal profile comparison of CD-1Y/CD-2Y versus maintained cervical alignment cases revealed no differences (P > .05) at any interval. Baseline C2-C7 sagittal vertical axis (SVA) was associated with increased new-onset CD risk at 1 (odds ratio [OR] 1.14, P = .025) and 2 years (OR 1.04, P = .032); prior spine surgical history was associated with CD risk at 1-year follow-up (OR 6.75, P = .047); baseline C2 slope was associated with increased CD risk at 2-year follow-up (OR 1.12, P = .041). CD development did not significantly impact health-related quality of life (P > .05).

Conclusions

Cervical deformity can manifest in nonoperative ASD patients: 30.0% at 1-year follow-up, and 41.7% at 2-year follow-up. Progressive CD manifested independently of thoracolumbar profile changes. Increased baseline C2-C7 SVA, C2 slope, and prior surgical history increased new-onset CD odds at 1 and 2 years.

Study design

Retrospective review of a multicenter prospective registry.

Objectives

Our goal was to develop a method to risk-stratify adult spinal deformity (ASD) patients on the basis of their accumulated health deficits. We developed a novel comorbidity score (CS) specific to patients with ASD based on their preoperative health state and investigated whether it was associated with major complications, length of hospital stay (LOS), and self-reported outcomes after ASD surgery.

Methods

We identified 273 operatively treated ASD patients with 2-year follow-up. We assessed associations between major complications and age, comorbidities, Charlson Comorbidity Index score, and Oswestry Disability Index score. Significant factors were used to construct the ASD-CS. Associations of ASD-CS with major complications, LOS, and patient-reported outcomes were analyzed.

Results

Major complications increased significantly with ASD-CS (P < .01). Compared with patients with ASD-CS of 0, the odds of major complications were 2.8-fold higher (P = .068) in patients with ASD-CS of 1 through 3; 4.5-fold higher (P < .01) in patients with ASD-CS of 4 through 6; and 7.5-fold higher (P < .01) in patients with ASD-CS of 7 or 8. Patients with ASD-CS of 7 or 8 had the longest mean LOS (10.7 days) and worst mean Scoliosis Research Society-22r total score at baseline; however, they experienced the greatest mean improvement (0.98 points) over 2 years.

Conclusions

The ASD-CS is significantly associated with major complications, LOS, and patient-reported outcomes in operatively treated ASD patients.

Study design

Retrospective cohort study.

Objective

Factors that predict outcomes for adult cervical spine deformity (ACSD) have not been well defined. To compare ACSD patients with best versus worst outcomes.

Methods

This study was based on a prospective, multicenter observational ACSD cohort. Best versus worst outcomes were compared based on Neck Disability Index (NDI), Neck Pain Numeric Rating Scale (NP-NRS), and modified Japanese Orthopaedic Association (mJOA) scores.

Results

Of 111 patients, 80 (72%) had minimum 1-year follow-up. For NDI, compared with best outcome patients (n = 28), worst outcome patients (n = 32) were more likely to have had a major complication (P = .004) and to have undergone a posterior-only procedure (P = .039), had greater Charlson Comorbidity Index (P = .009), and had worse postoperative C7-S1 sagittal vertical axis (SVA; P = .027). For NP-NRS, compared with best outcome patients (n = 26), worst outcome patients (n = 18) were younger (P = .045), had worse baseline NP-NRS (P = .034), and were more likely to have had a minor complication (P = .030). For the mJOA, compared with best outcome patients (n = 16), worst outcome patients (n = 18) were more likely to have had a major complication (P = .007) and to have a better baseline mJOA (P = .030). Multivariate models for NDI included posterior-only surgery (P = .006), major complication (P = .002), and postoperative C7-S1 SVA (P = .012); models for NP-NRS included baseline NP-NRS (P = .009), age (P = .017), and posterior-only surgery (P = .038); and models for mJOA included major complication (P = .008).

Conclusions

Factors distinguishing best and worst ACSD surgery outcomes included patient, surgical, and radiographic factors. These findings suggest areas that may warrant greater awareness to optimize patient counseling and outcomes.