Faculty Bibliography

OBJECTIVE:The substantial financial implications of minimally invasive surgery for adult spinal deformity (ASD) necessitate a thorough assessment of its inherent value and efficacy. Factors contributing to protracted cost-effectiveness (CE) have not been examined in the context of minimally invasive spine surgery (MIS) for ASD (MIS-ASD). Investigating these determinants can yield pivotal insight to optimize the efficacy of such surgical interventions while concurrently moderating associated expenditures. METHODS:MIS-ASD patients who underwent fusion > 2 levels with lateral lumbar interbody fusion or anterior lumbar interbody fusion and 4-year (4Y) follow-up were included. Published methods were used to determine the costs based on the Centers for Medicare & Medicaid Services definitions and average diagnosis-related group (DRG) reimbursement rates. Utility was calculated using quality-adjusted life years (QALYs), with a 3% discount applied for decline with life expectancy. Cost-utility (CU) was determined by dividing costs by total utility gained. Those who met CE at 4 years (CE4+) were evaluated relative to those who did not (CE4-). RESULTS:Eighty-six patients were included. Revision surgery occurred in 27% of patients. The overall mean cost was $73,000. CU at 4 years was $233,000, with 44% meeting CE4+ and a cumulative mean ± SD QALY gain of 0.8 ± 0.7. Among patients without revision, 54% met CE4+, while 76% met CE at life expectancy. There were no differences in length of stay, ICU admission rates, or time in ICU. Those with greater baseline disability (OR 1.1, p < 0.05) and frailty (OR 1.8, p < 0.05) had a higher likelihood of achieving CE4+. Lower comorbidity burden (i.e., lower Charlson Comorbidity Index score) was associated with increased odds of achieving CE4+ (OR 1.8, p < 0.05). Improved correction of pelvic incidence-lumbar lordosis mismatch was associated with achieving CE4+ (OR 3.8, p < 0.05). Those patients with major complications had 6× higher odds of failure to achieve CE4+, whereas those who underwent reoperation had 12× odds (both p < 0.05). CONCLUSIONS:MIS-ASD achieves CE in a significant subset of patients, particularly those with higher baseline disability, lower comorbidity burden, and better correction of spinal deformities. However, major complications and reoperations significantly hinder CE, underscoring the importance of optimizing patient selection and surgical techniques.

STUDY DESIGN/METHODS:Prospective multicenter study. OBJECTIVE:To determine the incidence of all-cause mortality after adult spinal deformity surgery. SUMMARY OF BACKGROUND DATA/BACKGROUND:Patients undergoing adult spinal deformity surgery are often frail and the procedures are invasive. The incidence of all-cause mortality among patients undergoing cervical or thoracolumbar deformity surgery is unclear. METHODS:Using 2 prospective, multicenter databases, we identified patients who underwent surgery for cervical deformity surgery from 2013-2020 (n=169) or thoracolumbar deformity from 2008-2020 (n=1507). Mortality incidence density was calculated as follows: 100 × (number of deaths) / (sum of total years of follow-up for all patients). RESULTS:Of 169 participants in the cervical group (mean±standard deviation age, 61±10 y), death occurred in 19 (11%). The mean time to death was 25±19 months. Mortality incidence density was 4.4 deaths per 100 person-years. The 30-day mortality rate was 0.6% (1/169) and 90-day mortality rate was 1.2% (2/169). The 3 most common causes of death were arrhythmia/cardiac arrest (16%), congestive heart failure (11%), and pneumonia (11%). There were no intraoperative deaths. Of 1507 participants in the thoracolumbar group (mean±standard deviation age, 61±14 y), death occurred in 53 (3.5%). The mean time to death was 32.5±21.5 months. Mean duration of follow-up was 1.8±1.5 years. The mortality incidence density was 0.8 deaths per 100 person-years. The 30-day mortality rate was 0.1% (1/1507) and 90-day mortality rate was 0.3% (4/1507). The 3 most common causes of death were non-spine malignancy (13%), pneumonia (9%), and arrhythmia/cardiac arrest (6%). CONCLUSIONS:The number of deaths per year was higher among cervical deformity patients (4.4 per 100 person-years) than among thoracolumbar deformity patients (0.8 per 100 person-years). Pneumonia and arrhythmia/cardiac arrest were among the most common causes of death in both groups. LEVEL OF EVIDENCE/METHODS:III.

STUDY DESIGN/METHODS:Prospective, multicenter study. OBJECTIVE:To compare and characterize complications in adult spinal deformity (ASD) patients with and without 3-column osteotomy (3CO). SUMMARY OF BACKGROUND DATA/BACKGROUND:Although 3CO is associated with increased risk of neurologic adverse events, no study has, to our knowledge, compared and characterized sensory and motor neurologic complications in ASD patients with vs. without 3CO. MATERIALS AND METHODS/METHODS:Demographics, surgical characteristics, and neurologic complications were collected for 553 ASD patients. Lower extremity motor scores (LEMSs) were compared at baseline and postoperatively. Multivariate analysis was performed to identify risk factors associated with neurologic adverse events. RESULTS:Among 553 ASD patients, 130 (23.5%) underwent 3CO. More patients with 3CO were revision patients (67.7% vs. 35.0%; P<0.001), presented with sagittal deformity (43.9% vs. 31.0%; P=0.008), and had longer operative times (455.6 vs. 407.3 min; P=0.001) and greater estimated blood loss (EBL) (1,650 vs. 1,000 mL; P<0.001) than patients without 3CO. The incidence of neurologic adverse events was greater among patients with vs. without 3CO (23.1% vs. 15.4%; P=0.04). Multivariate analysis revealed that older age (odds ratio [OR] 1.288 per 10-year increase, 95% confidence interval [CI] 1.060-1.565, P=0.01) and longer operative time (OR 1.088, 95% CI 1.001-1.004, P=0.01) were significant predictors of neurologic adverse events. No between-group difference in LEMS was observed at 6-week (49.0 vs. 49.0; P=0.90) or 1-year (49.4 vs. 49.3; P=0.71) follow-up. By 1-year follow-up, 20.5% of 3CO patients and 21.6% of patients without 3CO had residual motor deficit. CONCLUSIONS:Compared with patients who did not undergo 3CO, more patients with 3CO had prior instrumentation, presented with sagittal deformity, had longer operative time and greater EBL, and were more likely to have a neurologic adverse event. At 1-year follow-up, there was no significant difference in LEMS between the two groups. LEVEL OF EVIDENCE/METHODS:III.

OBJECTIVE:The all-payer model is a healthcare payment system unique to the state of Maryland, while the Medicare Severity-Diagnosis-Related Group (MS-DRG) model is used by all other states. The purpose of this study was to investigate differences in reimbursement and inpatient length of stay (LOS) in adult spinal deformity (ASD) surgery between the all-payer and MS-DRG models. METHODS:MS-DRG reimbursements were calculated using the Centers for Medicare & Medicaid Services Web Pricer tool; reimbursements for the all-payer model were compiled from a single institution in the state of Maryland. Payments for the most frequently occurring ASD MS-DRGs (codes 453, 454, 455, 456, 457, 458, and 460) were analyzed for fiscal years 2018-2023. The mean inpatient LOS was calculated for each MS-DRG code and reimbursement model. RESULTS:When comparing 416 MS-DRG and 1783 all-payer model accounts, the overall mean reimbursements for ASD surgery were significantly lower under the MS-DRG model ($59,199 vs $77,246, p < 0.001). The mean reimbursement payments for MS-DRG codes 453, 454, 455, 456, 457, and 460 were significantly higher under the Maryland all-payer model (p < 0.001). The mean LOS was significantly shorter in the all-payer model for MS-DRG code 453 (p = 0.046) and longer for code 457 (p < 0.001). For all other codes, no significant differences in LOS were observed. CONCLUSIONS:ASD surgery reimbursements are higher overall under the Maryland all-payer model compared with the MS-DRG model. The mean inpatient LOS did not differ significantly across most MS-DRG codes, highlighting the financial viability of an all-payer model in a healthcare system.

STUDY DESIGN/METHODS:Prospective, multi-center analysis. OBJECTIVE:Evaluate preoperative and minimum 2-year postoperative health related quality of life (HRQOL) outcomes for adult spine deformity (ASD) using legacy HRQOL measures and computer adaptive testing (CAT) version of the patient reported outcome measurement information system (PROMIS). SUMMARY OF BACKGROUND DATA/BACKGROUND:PROMIS is the NIH recommended measure for patient reported outcomes. No data exists evaluating minimum 2-year ASD surgical outcomes using PROMIS CAT compared to legacy HRQOL measures. METHODS:ASD patients > 18 years of age were enrolled into a prospective, multi-center, observational study. Patients were administered legacy HRQOLs (SRS-22r, ODI, NRS back and leg pain, VR-12) and PROMIS CAT for Pain Interference, Physical Function, Satisfaction with Participation in Social Roles, Satisfaction with Participation in Discretionary Social Activities, Depression, and Anxiety. Preoperative and minimum 2-year postoperative HRQOL scores were compared and ASD PROMIS scores were compared to general population PROMIS values. RESULTS:144 of 190 patients had complete data at minimum 2-year follow-up. Mean preoperative values included: age=62.6 years, scoliosis=34.2°, SVA=74.2mm, PI-LL=16.7. At mean 2.3 year follow-up surgical treatment improved spine alignment (scoliosis=14.7°, SVA=34.5mm, and PI-LL=4.5°; P<0.0001). All legacy HRQOL and PROMIS scores improved from preoperative to minimum 2-year postoperative (P<0.005). Referencing PROMIS scores to normative values demonstrated 50-85% of ASD patients preoperatively reported moderate-severe deficits in all domains. At minimum 2-year postoperative 52-88% of ASD patients reported normal values-mild deficits in all PROMIS domains. PROMIS social health domains demonstrated the greatest improvement. CONCLUSION/CONCLUSIONS:This is the first study evaluating minimum 2-year ASD surgical outcomes using PROMIS CAT. ASD patients demonstrated postoperative improvements in all PROMIS health domains, with the social health domains improving most. The majority of ASD patients reported normal PROMIS values at minimum 2-years postoperative.

STUDY DESIGN/METHODS:Retrospective review of multicenter data. OBJECTIVE:To compare long-term neurologic recovery in patients with and without incidental durotomy (hereafter, "durotomy") after adult spinal deformity surgery. SUMMARY OF BACKGROUND DATA/BACKGROUND:Durotomy is a common complication of adult spinal deformity surgery and is typically associated with technical challenges during the procedure. METHODS:Using a prospectively collected database, we included 1452 patients (73% female; mean age, 60±14 y) who underwent adult spinal deformity surgery from 2008-2020 at 22 US centers. We compared patients with and without durotomy with respect to demographic characteristics, surgical variables, and neurologic outcomes at baseline and at 1 and 2 years postoperatively. Multivariate analysis compared neurologic complications and length of stay (LOS) between the groups. P<.05 was considered significant. RESULTS:Durotomy occurred in 121 patients (8.3%). Patients with durotomy were more likely to have undergone revision surgery (P<.001) and had higher Charlson Comorbidity Index values (P=.029) than those who did not. Patients with durotomy had higher estimated blood loss, longer operative time, more frequent 3-column osteotomies, and longer LOS (all, P<.001). Lower-extremity motor scores did not differ between patients with durotomy and those without at 1 and 2 years postoperatively. The incidence of neurologic, medical, and surgical complications did not differ significantly between the 2 groups. Patients with durotomy had a higher rate of inpatient return to the operating room (5.0%) than those without (2.0%) (P=.04). On multivariate analysis, there were no differences between groups in lower-extremity motor scores, neurologic complications, or LOS. CONCLUSIONS:Incidental durotomy during adult spinal deformity surgery was associated with greater intraoperative complexity and transient sensory symptoms but did not adversely affect long-term motor recovery, neurologic complications, or patient-reported outcomes. These findings suggest durotomy is a manageable complication without lasting functional consequences.

STUDY DESIGN/METHODS:Retrospective review of prospectively collected, multicenter data. OBJECTIVE:To assess associations between patient height and weight independently and interactively with the incidence of proximal junctional kyphosis (PJK) after surgical treatment of adult spine deformity. SUMMARY OF BACKGROUND DATA/BACKGROUND:Body mass index has traditionally been used to assess the influence of body composition on surgical outcomes, but the individual effects of height and weight have not been studied in relation to PJK. METHODS:We compared baseline demographic characteristics, radiographic measurements, and perioperative variables between patients who developed PJK after adult spinal deformity surgery between 2008 and 2020 and those who did not. Using a generalized additive model with a logistic link function, we modeled height and weight and their interaction as smooth terms to capture potential nonlinear effects on PJK risk. Multivariate analysis was adjusted for age, history of osteoporosis, upper instrumented vertebra, number of levels fused, and postoperative pelvic incidence minus lumbar lordosis and T1 pelvic angle. RESULTS:Of 904 included patients, the median age was 65 years (interquartile range: 58-71), and 76% were female. PJK developed in 131 patients (14%). Baseline characteristics, including frailty, comorbidities, and radiographic measures, did not differ significantly between the PJK and non-PJK groups. Taller height was a predictor of PJK (P=.03). In contrast, weight was not an independent predictor, and there was no significant interaction between height and weight. The incidence of PJK peaked at a height of approximately 179 cm before plateauing. CONCLUSIONS:Taller height, but not weight, was associated with developing PJK after adult spinal deformity surgery. These findings underscore the importance of considering patient height during surgical planning.

STUDY DESIGN/METHODS:Retrospective Multi-Center Study. OBJECTIVE:To investigate how advances in spine realignment have impacted lumbar segmental alignment. SUMMARY OF BACKGROUND DATA/BACKGROUND:The understanding of spine alignment and Adult Spinal Deformity (ASD) management continues to advance. It remains unknown how these advances have influenced lumbar segmental alignment changes in the setting of surgical correction. METHODS:Patients undergoing primary thoracolumbar fusion for ASD were stratified based on enrolment in two distinct multicenter registries; forming an 'Early' (2008-2017) and a 'Late cohort' (2018-present). Patients were further stratified based on pelvic incidence (PI) and Roussouly type. Segmental alignment was determined based on published values of asymptomatic individuals. Pelvic incidence-based alignment and Roussouly-based alignment were determined in alignment with previously published normative values. Means comparisons tests and multivariate analyses compared segmental & regional parameters between groups. RESULTS:1240 patients included (622 EARLY, 618 LATE). The mean age was 61.4±14.5 years, body mass index (BMI) was 28.0±5.8 kg/m2, and Charlson comorbidity index (CCI) was 1.55±1.70. 70.2% of patients were female gender. LATE consistently displayed better L5-S1 alignment across all PI and Roussouly types (P=0.001) However, EARLY demonstrated better L4-5 alignment (P=0.001). Improved alignment in L5-S1, L4-5 and L3-4 was associated with achieving minimum clinically important difference in ODI scores and decreased risk of mechanical complications. Both cohorts demonstrated low rates of matching L4-S1 regional and overall lumbar lordosis L1-S1 alignment, with no differences between both groups. By lordosis distribution index, both groups had predominantly hyperlordotic maldistribution postop, but LATE had more 'Aligned' patients (15.9 vs. 11%, P<0.001). CONCLUSIONS:Over the past 15 years, surgeons appear to be better at restoring ideal lumbar segmental sagittal alignment in ASD patients. However, idealized correction does not appear to be uniform across all lumbar segments, representing an opportunity for further improvement. LEVEL OF EVIDENCE/METHODS:III.

STUDY DESIGN/METHODS:Retrospective cohort study. OBJECTIVE:To examine the degree of overlap between disability, multimorbidity, and frailty in a cohort of ASD patients. SUMMARY OF BACKGROUND DATA/BACKGROUND:Frailty is a popular topic in spine research, as it is a reported risk factor for poor outcomes. Disability, multimorbidity, and frailty can coexist, sometimes causing or exacerbating one another. It is important to distinguish these conditions for perioperative optimization and to guide research initiatives. METHODS:A multicenter registry of ASD patients was queried for baseline data regarding frailty, as measured by the Edmonton Frail Scale, disability, as measured by the Oswestry Disability Index, and multimorbidity, as measured by the Charlson Comorbidity Index. The relationships between these measures and both chronological and biological age (PhenoAge) were explored. Exploratory factor analysis (EFA) examined areas of overlap between these diagnoses. RESULTS:There were 861 patients contributing data, mostly female (68%), most undergoing primary surgery at a median age of 66 years (Interquartile Range (55.1-71.6), with 6% classified as "Frail." Chronological and PhenoAge showed weak to moderate associations with disability and frailty, though PhenoAge was stronger. There was no evidence of distinct clusters, rather a continuity of condition severity. EFA found overlap between subjective and objective measures of disability, function, and frailty. CONCLUSIONS:Frailty was rare (6%) in this multicenter cohort of patients. Conflation of disability and frailty is a real risk due to overlap in measures of both conditions. Disability and frailty do not form discrete categories but rather exist along a continuum, underscoring the need to abandon categorical labels in favor of continuous measures for both clinical assessment and research settings.

STUDY DESIGN/METHODS:Retrospective cohort analysis. OBJECTIVE:Assess the distribution of Medicare's spinal-deformity-specific diagnosis-related group (DRGs) relative to surgical invasiveness, hospital length of stay (LOS), discharge disposition, 90-day postoperative quality metrics, and reimbursement costs for adult spinal deformity (ASD) operations. SUMMARY OF BACKGROUND DATA/BACKGROUND:Heterogeneity of ASD call into question Medicare's DRGs to accurately capture nuances of ASD surgical episodes of care. METHODS:Adults who underwent thoracic to pelvis instrumentation with associated DRGs were identified from a multicenter database. Demographics, operative details, inpatient course, discharge disposition, 90-day adverse events, and reimbursement costs were compared between spinal deformity-specific DRG codes. Distribution of DRGs for a subset of these patients who fit into one of 6 commonly performed surgical strategies to address ASD was also assessed. RESULTS:Of the 314 patients included for analysis, the majority fell into +CC DRGs, while the minority had +MCC DRGs or no MCC/CC DRG. Within each DRG, there was considerable heterogeneity in regard to patients' ages, ASA, CCI, frailty, surgical invasiveness, postoperative ICU/hospital LOS, discharge disposition, and complication profiles.+MCC DRGs had significantly greater ASA and Edmonton Frailty Scores. While +MCC and +CC had relatively similar surgical invasiveness, +MCC had greater ICU admissions, in-hospital adverse events, and nonhome discharges as well as longer ICU, hospital, and rehab LOS. While reimbursements were significantly higher for +MCC DRG compared with +CC DRGs and DRGs without MCC/CC, there were large ranges in reimbursement within all DRG subgroups.The 7 DRGs varied significantly within and between the subset of 6 commonly performed surgical strategies, although there were no differences in regard to ICU admissions and LOS, hospital LOS, discharge disposition, and number of adverse events (in-hospital, 90-day). CONCLUSIONS:While Medicare's spinal-deformity DRG codes capture average trends in surgical/postoperative episodes of care for ASD patients, each encompasses highly heterogeneous patients and associated surgical operations rendering them unreliable gauges of patient/surgical complexity, early postoperative trajectories, and reimbursement costs. A more granular system is needed to more accurately capture the nuances of ASD operations and their associated quality metrics and reimbursement costs.