Faculty Bibliography

Introduction: Blood loss is a major concern in spine surgery. Blood transfusion promotes oxygen delivery and tissue perfusion during long, complex surgeries, yet carries with it rare but notable risks. The hemoglobin (Hb) trigger-the Hb value that initiates clinician administration of packed red blood cells (PRBCs)-is frequently used to evaluate physician compliance with existing transfusion guidelines. Randomized clinical trials have demonstrated similar or improved outcomes among patients receiving blood transfusions using a restrictive Hb trigger-defined as an intraoperative Hb level of <10 g/dL intraoperatively or <8 g/dL postoperatively-versus a liberal Hb trigger (>=10 g/dL intraoperatively or >=8 g/dL postoperatively) in cardiac and hip surgery. To the authors' knowledge, no study has examined the associated morbidity and financial impact of liberal transfusions within spinal surgery. We thus aimed to determine the perioperative clinical outcomes and costs associated with liberal versus restrictive transfusion triggers among spine surgery patients. Material and Methods: The surgical billing database at our institution was queried for inpatients discharged following spinal surgery between 2008 and 2015, yielding 33,043 patients. Patients were stratified into eight groups according to the spine surgical procedure performed. The values used were the reported institutional acquisition cost ($220/unit) and a mean estimated activity-based cost ($760/unit) based on a Society for the Advancement of Blood Management report. Outcomes considered included mortality, in-hospital morbidity, total costs, and length of stay. Results: A total of 6,931 patients met the inclusion criteria. PRBC transfusions occured in 2,374 patients (at least 1 unit of PRBCs), yielding an overall transfusion rate of 34.3%. Compared to the other surgical groups, most PRBC use occurred within posterior lumbar fusion patients. The mean intraoperative Hb trigger was 10.1 (SD = 1.7) g/dL and the mean postoperative Hb trigger was 10.1 (SD = 2.1) g/dL. For the 2,374 patients that were transfused with PRBCs, 1420 (59.8%) received a postoperative PRBC transfusion with a liberal postoperative Hb trigger (>=8 g/dL) while 529 (22.3%) received an intraoperative PRBC transfusion with a liberal intraoperative Hb trigger (>=10 g/dL). Logistic regression analysis revealed that patients with a nadir Hb of 8-10 g/dL transfused with PRBC had an independently higher risk of perioperative morbidity (odds ratio [OR] = 2.12; 95% confidence interval [CI], 1.24-3.64; P = .006). The additional cost when comparing restrictive and liberal transfusion triggers was estimated to be between $1,330,439 and $4,596,062 (58.7% of the estimated total cost of PRBC transfusion), with the cost varying by surgery type. When comparing liberal and restrictive PRBC transfusion triggers, an estimated additional $202,675 to $700,151 in institutional costs were incurred each year among patients undergoing spine surgery. Conclusion: Patients with a nadir Hb of 8 to 10 g/dL who were transfused had higher perioperative morbidity, even after adjusting for potential confounders. The additional cost incurred from liberal transfusion trigger reliance ranged from an estimated $202,675 to $700,151 annually. These findings point to a potential area for clinicians and institutions to improve patient outcomes and reduce costs

Introduction: Despite the potential for profound impact of adult cervical deformity (ACD) on function and healthrelated quality of life, there remains a paucity of highquality studies that assess outcomes of surgical treatment for these patients. Our objective was to assess outcomes following surgical treatment for ACD based on a prospective multicenter consecutive case series. Materials and Methods: Surgically treated ACD patients eligible for 1-year followup were identified from a prospectively collected multicenter database. Baseline deformity characteristics, surgical parameters, and 1-year outcomes were assessed. Standardized outcome measures included: Neck Disability Index (NDI, range 0-100), neck pain numeric rating scale (NRS) score (range 0-10), and EQ-5D index (range 0 -1) and subscores (range 1-3). Paired sample t-tests were used to compare 1-year and baseline measures. Results: Of 77 ACD patients, 55 (71%) had 1-year follow-up (64% women, mean age 61 years, mean Charlson Comorbidity Index [CCI] of 0.6, previous cervical surgery in 44%). Diagnoses included: cervical sagittal imbalance (62%), cervical kyphosis (60%), proximal junctional kyphosis (8%), and coronal deformity (10%). Posterior fusion was performed in 85% (mean number of vertebral levels=10), and anterior fusion was performed in 29% (mean number of vertebral levels = 5). Three-column osteotomy was performed in 24% of patients. Mean operative time was 6.5 hours and mean estimated blood loss was 0.9L. At 1-year following surgery, ACD patients had significant improvement in NDI (50.5 to 38.0, P < .001), neck pain NRS (6.9 to 4.3, P < .001), EQ-5D index (0.51 to 0.66, P < .001), and EQ-5D subscores: mobility (1.9 to 1.7, P = .019), usual activities (2.2 to 1.9, P = .007), pain/discomfort (2.4 to 2.1, P < .001), anxiety/depression (1.8 to 1.5, P = .014). A nonsignificant trend favoring improvement was observed for EQ-5D self-care (1.5 to 1.3, P = .070). Compared with patients that achieved 1-year follow-up, those lost to followup did not differ significantly with regard to age, gender, CCI, number of fused anterior or posterior vertebral levels, or baseline NDI, neck pain NRS, or EQ-5D scores. Conclusions: Based on a prospective multicenter series of adults with cervical deformity, surgical treatment provided significant improvement in multiple measures of pain and function, including the NDI, neck pain NRS score, and EQ-5D. Further follow-up will be necessary to assess the durability of these surgical procedures and the resulting improved outcomes

Introduction: Although adult cervical deformity (ACD) has been empirically associated with significant pain and disability, the magnitude of this negative impact has not been objectively quantified. Our objective was to assess whether symptomatic ACD patients have substantial negative health impact based on the EQ-5D compared with United States (U.S.) normative and chronic disease state values. Materials and Methods: ACD patients presenting for surgical evaluation and treatment were identified from a prospectively collected multicenter database. Baseline demographics, deformity characteristics, and EQ-5D scores were collected. EQ-5D scores were compared with ageand gender-matched U.S. normative and chronic disease state values. Results: Of 121 ACD patients, 115 (95%) completed the EQ-5D (61% women, mean age 61 years, previous cervical surgery in 46%). Diagnoses included: cervical sagittal malalignment (63%), cervical kyphosis (60%), proximal junctional kyphosis (9%) and coronal deformity (8%). The mean EQ-5D index was 0.511, which is 35% below the bottom 25th percentile score (0.790) for a similar age- and gender-weighted normative population and worse than the bottom 25th percentile for several other chronic disease states (diabetes [0.708], ischemic heart disease [0.708], and myocardial infarction [0.575]). The EQ-5D index of 0.511 seen in this ACD cohort is comparable to the bottom 25th percentile for blindness (0.543), emphysema (0.508) and heart failure (0.437). Based on EQ-5D subscores, patients reported impact on mobility (87%), daily self-care (47%), daily activities (91%), pain/discomfort (98%), and anxiety/depression (67%). Conclusions: The health impact of symptomatic ACD is substantial, with an EQ-5D index that is 35% below the bottom 25th percentile for an age- and gender-matched normative U.S. population. The markedly negative health impact of ACD was demonstrated across all domains of the EQ-5D. The overall mean EQ-5D index for ACD patients was worse than the bottom 25th percentile for several other chronic disease states, including chronic ischemic heart disease, malignant breast cancer, and malignant prostate cancer, and was comparable to the bottom 25th percentile values for blindness/low vision, emphysema, renal failure, and stroke

Introduction: Although three-column osteotomy (3CO; pedicle subtraction osteotomy [PSO] or vertebral column resection [VCR]) can provide powerful alignment correction and disability improvement in adult cervical deformity (ACD), these procedures are complex and associated with high complication rates. Previous reports on complications associated with 3CO for ACD have been primarily based on retrospective complication collection, which may substantially underestimate the true rates. Our objective was to provide a prospective assessment of cervical alignment improvement and complications in ACD patients treated with 3CO. Materials and Methods: Although three-column osteotomy (3CO; pedicle subtraction osteotomy [PSO] or vertebral column resection [VCR]) can provide powerful alignment correction and disability improvement in adult cervical deformity (ACD), these procedures are complex and associated with high complication rates. Previous reports on complications associated with 3CO for ACD have been primarily based on retrospective complication collection, which may substantially underestimate the true rates. Our objective was to provide a prospective assessment of cervical alignment improvement and complications in ACD patients treated with 3CO. Results: All 24 ACD patients treated with 3CO (15 PSO/ 9 VCR) achieved minimum 90-day follow-up (71% women, mean age 62 years, previous surgery in 54%). Diagnoses included: cervical sagittal imbalance (92%), cervical kyphosis (38%), proximal junctional kyphosis (17%), coronal deformity (8%) and distal junctional kyphosis (4%). The mean number of posterior fusion levels was 13, and 4% also had an anterior fusion. The most common 3CO levels were T1 (38%), T2 (29%) and T3 (21%). A total of 25 (19 major/6 minor) complications were reported, with 14 (58%) and 6 (25%) patients affected, respectively. Overall, 17 (71%) patients had at least one complication. The most common complications were excessive blood loss (>1.7L, 25%), neurologic deficit (17%), distal junctional kyphosis (DJK, 8%), wound infection (13%), and cardiorespiratory failure (8%). Four (17%) patients required re-operation within 90-days (2 for nerve root motor deficit, 1 deep wound infection, 1 implant pain/prominence). Cervical sagittal alignment improved significantly following 3CO: cervical lordosis (CL, 3degree to 13degree, P = .031), C2-7 sagittal vertical axis (66mm to 44mm, P < .001), and T1 slope minus CL (46degree to 27degree, P < .001). Conclusions: Among 24 ACD patients treated with 3CO, cervical sagittal alignment improved significantly following surgery. Overall, 17 (71%) patients had at least one complication (19 major/6 minor). The most common complications were excessive blood loss (>1.7L), neurologic deficit, DJK, wound infection, and cardiorespiratory failure. Future research focused on reducing these complications may present the greatest opportunities for safety and cost improvements for these procedures

Introduction: The length of stay (LOS) following adult spinal deformity (ASD) surgery is a critical time period allowing for recovery to levels safe enough for to return home or to rehabilitation. Thus, the goal is to minimize it for conserving hospital resources and third party payer pressure. Factors related to LOS have not been studied nor has a predictive model been created. The purpose was to create a preoperative predictive model to predict the LOS following ASD surgery. Material and Methods: Retrospective review of a multicenter, prospective ASD database. Inclusion criteria: operative pts, age >18 yrs, ASD. Pts with staged surgery at a separate hospitalization or LOS >30 days were excluded. 66 variables were initially evaluated with 40 being used for model building following univariable predictor importance >=0.90, redundancy, and collinearity testing. Variables included: demographic data, comorbidities, preop HRQOL, preop coronal and sagittal radiographic parameters, and modifiable surgical factors. A generalized linear model was constructed using a training dataset developed from a boostrapped sample with replacement using a random number generator. Pts randomly omitted from the boostrapped sample were included in the testing dataset. Accuracy was calculated by comparison of predicted LOS to the actual LOS. Results: A total of 689 patients were eligible with 653 meeting inclusion criteria. The mean LOS was 7.9 +/- 4.1 days (median = 7, range: 1-28). Following bootstrapping, a total of 893 pts were modeled, Training: 653, Testing: 240(36.6%). The linear correlations for the training and testing datasets were 0.632 and 0.507, respectively. Testing dataset accuracy within 2 days of actual LOS was 75.4% (181/240 pts). The top 10 predictors were the following in decreasing order: staged surgery (yes/ no), C7 SVA, number of posterior levels fused, Charlson Comorbidity Index, Total number of comorbidities, preop ODI score, iliac fixation (yes/no), preop SRS Activity score, preop SRS Appearance scores, and number of interbody fusion levels. Conclusion: A successful model was created to predict LOS following ASD surgery to an accuracy of 75% within 2 days. There are some factors related to LOS that are not likely captured in large databases, which may partially explain the 75% accuracy, such as rehab bed availability and social support resources

BACKGROUND: Recent studies show increases in cervical spine surgery prevalence and cervical spondylotic myelopathy (CSM) diagnoses in the US. However, few studies have examined outcomes for CSM surgical management, particularly on a nationwide scale. OBJECTIVE: Evaluate national trends from 2001 to 2010 for CSM patient surgical approach, postoperative outcomes, and hospital characteristics. METHODS: A retrospective nationwide database analysis provided by the Nationwide Inpatient Sample (NIS) including CSM patients aged 25+ who underwent anterior and/or posterior cervical fusion or laminoplasty from 2001 to 2010. Patients with fractures, 9+ levels fused, or any cancer were excluded. Measures included demographics, hospital data, and procedure-related complications. Yearly trends were analyzed using linear regression modeling. RESULTS: 54,348 discharge cases were identified. ACDF, posterior only, and combined anterior/posterior approach volumes significantly increased from 2001 to 2010 (98.62%, 303.07%, and 576.19%; respectively, p<0.05). However, laminoplasty volume remained unchanged (p>0.05). Total charges for ACDF, posterior only, combined anterior/posterior, and laminoplasty approaches all significantly increased (138.72%, 176.74%, 182.48%, and 144.85%, respectively; p<0.05). For all procedures, overall mortality significantly decreased by 45.34% (p=0.001) and overall morbidity increased by 33.82% (p=0.0002). For all procedures except ACDF, which saw a significantly decrease by 8.75% (p<0.0001), length of hospital stay was unchanged. CONCLUSIONS: For CSM patients between 2001 and 2010, combined surgical approach increased sixfold, posterior only approach increased threefold, and ACDF doubled; laminoplasties without fusion volume remained the same. Mortality decreased whereas morbidity and total charges increased. Length of stay decreased only for ACDF approach. This study provides clinically useful data to direct future research, improving patient outcomes.

OBJECT: We sought to create a model capable of predicting the magnitude of pelvic incidence-lumbar lordosis (PI-LL) correction necessary to achieve a desired change in sagittal vertical axis (SVA). METHODS: Retrospective review of a prospectively maintained multicenter adult spinal deformity database collected by the International Spine Study Group between 2009 and 2014. The independent variable of interest was the degree of correction achieved in the PI-LL mismatch at 6-weeks post-surgery. Primary outcome was the change in global sagittal alignment at 6 weeks and 1 year after surgery. We used a linear mixed-effects model to determine the extent to which corrections in the PI-LL relationship affected post-operative changes in SVA. RESULTS: A total of 1053 adult patients were identified. Of these, 590 were managed surgically. 87 surgically managed patients were excluded due to incomplete or missing PI-LL measurements on follow-up; the remaining 503 of these patients were selected for inclusion. For each degree of improvement in the PI-LL mismatch at 6 weeks, the SVA decreased by 2.18 mm (95% CI:-2.56, -1.79, p<0.01) and 1.67 mm (95% CI:-2.07, - 1.27, p<0.01) at 6 weeks and 12 months, respectively. A high SVA measurement (>50 mm) at 1-year post-surgery was negatively associated with health-related quality of life as measured by the SRS-12 outcomes assessment. CONCLUSION: We describe a novel model that illustrates how surgical correction of the PI-LL relationship affects post-operative changes in SVA. This may enable surgeons to determine pre-operatively the amount of lumbar lordosis necessary to achieve a desired change in SVA.

BACKGROUND: Although adult cervical spine deformity (ACSD) is associated with pain and disability, its health impact has not been quantified in comparison to other chronic diseases. OBJECTIVE: To perform a comparative analysis of the health impact of symptomatic ACSD to US normative and chronic disease values using EQ-5D (EuroQuol-5 Dimensions questionnaire) scores. METHODS: ACSD patients presenting for surgical treatment were identified from a prospectively collected multicenter database. Baseline demographics and EQ-5D scores were collected and compared with US normative and disease state values. RESULTS: Of 121 ACSD patients, 115 (95%) completed the EQ-5D (60% women, mean age 61 years, previous spine surgery in 44%). Diagnoses included kyphosis with mid-cervical (63.4%), cervico-thoracic (23.5%), or thoracic (8.7%) apex and primary coronal deformity (4.3%). The mean ACSD EQ-5D index was 0.511 (standard definition = 0.224), which is 34% below the bottom 25th percentile (0.780) for similar age- and gender-matched US normative populations. Mean ACSD EQ-5D index values were worse than the bottom 25th percentile for several other disease states, including chronic ischemic heart disease (0.708), malignant breast cancer (0.708), and malignant prostate cancer (0.708). ACSD mean index values were comparable to the bottom 25th percentile values for blindness/low vision (0.543), emphysema (0.508), renal failure (0.506), and stroke (0.463). EQ-5D scores did not significantly differ based on cervical deformity type ( P = .66). CONCLUSION: The health impact of symptomatic ACSD is substantial, with negative impact across all EQ-5D domains. The mean ACSD EQ-5D index was comparable to the bottom 25th percentile values for blindness/low vision, emphysema, renal failure, and stroke.

BACKGROUND CONTEXT: Complication rates for adult cervical deformity are poorly characterized given the complexity and heterogeneity of cases. PURPOSE: To compare peri-operative complication rates following adult cervical deformity corrective surgery between a prospective multi-center database for cervical deformity patients (PCD) and Nationwide Inpatient Sample (NIS). STUDY DESIGN/SETTING: Retrospective review of prospective databases. PATIENT SAMPLE: 11,501 adult cervical deformity patients (11,379 patients from NIS and 122 patients from PCD database). OUTCOME MEASURES: Peri-operative medical and surgical complications. METHODS: The NIS was queried (2001-2013) for cervical deformity discharges for patients >/=18yrs undergoing cervical fusions using ICD-9 coding. Patients >/=18 years from the PCD database (2013-2015) were selected. Equivalent complications were identified and rates were compared. Bonferroni correction (p<0.004) was used for Pearson chi2. Binary logistic regression was used to evaluate differences in complication rates between databases. RESULTS: 11,379 NIS patients and 122 PCD patients were identified. PCD patients were older (62.49 vs. 55.15,p<0.001) but displayed similar gender distribution. Intra-operative complication rate was higher in PCD (39.3%) compared to NIS (9.2%,p<0.001). PCD had an increased risk of reporting overall complications than NIS (OR:2.81, CI:1.81-4.38). Only device-related complications were greater in NIS (7.1% vs. 1.1%,p=0.007). PCD patients displayed higher rates of the following complications: peripheral vascular (0.8% vs. 0.1%,p=0.001), GI (2.5% vs. 0.2%,p<0.001), infection (8.2% vs. 0.5%,p<0.001), dural tear (4.1% vs. 0.6%, p<0.001), and dysphagia (9.8% vs. 1.9%,p<0.001). GU, wound, and DVT complications were similar between databases (p>0.004). Based on surgical approach, PCD reported higher GI and neurologic complication rates for combined anterior/posterior procedures (p<0.001). For posterior-only procedures, NIS had more device-related complications (12.4% vs. 0.1%,p=0.003), while PCD had more infections (9.3% vs. 0.7%,p<0.001). CONCLUSIONS: Analysis of the surgeon-maintained cervical database revealed higher overall and individual complication rates and higher data granularity. The nationwide database may underestimate ACD patient complications particularly in regards to peri-operative surgical details due to coding and deformity generalizations. The surgeon-maintained database captures the surgical details, but may underestimate some medical complications.

BACKGROUND CONTEXT: Blood transfusions in spine surgery are shown to be associated with increased patient morbidity. The association between transfusion performed using a liberal hemoglobin trigger-defined as an intraoperative hemoglobin level of >/=10 g/dL, a postoperative level of >/=8 g/dL, or a whole hospital nadir between 8-10 g/dL-and perioperative morbidity and cost in spine surgery patients is unknown and thus was investigated in this study. PURPOSE: To describe the perioperative outcomes and economic cost associated with liberal hemoglobin trigger transfusion among spine surgery patients. STUDY DESIGN/SETTING: Retrospective study. PATIENT SAMPLE: The surgical billing database at our institution was queried for inpatients discharged between 2008 and 2015 after the following procedures: atlantoaxial fusion, anterior cervical fusion, posterior cervical fusion, anterior lumbar fusion, posterior lumbar fusion, lateral lumbar fusion, other procedures, and tumor-related surgeries. In total, 6931 patients were included for analysis. OUTCOME MEASURES: The primary outcome was composite morbidity which was comprised of: (1) infection (sepsis, surgical-site infection, Clostridium dificile, or drug-resistant infection), (2) thrombotic event (pulmonary embolus, deep venous thrombosis, or disseminated intravascular coagulation), (3) kidney injury, (4) respiratory event, and/or (5) ischemic event (transient ischemic attack, myocardial infarction, or cerebrovascular accident). METHODS: Data on intraoperative transfusion were obtained from an automated prospectively collected anesthesia data management system. Data on postoperative hospital transfusion was obtained through a Web-based intelligence portal. Based on previous research, we analyzed the data using three definitions of a liberal transfusion trigger in patients who received an RBC transfusion: a liberal intraoperative Hb trigger as a nadir Hb level of 10 g/dL or greater, a liberal postoperative Hb trigger as a nadir Hb level of 8 g/dL or greater, or a whole hospital nadir Hb level of 8-10 g/dL. Variables analyzed included in-hospital morbidity, mortality, length of stay, and total costs associated with a liberal transfusion strategy. RESULTS: Among patients with a whole hospital stay nadir hemoglobin between 8 to 10 g/dL, transfused patients demonstrated a longer in-hospital stay (median [IQR], 6 [5-9] vs. 4 [3-6] days; P<0.0001) and a higher perioperative morbidity (n=145, [11.5%] vs. n=74, [6.1%]; P<0.0001) than those not transfused. Even after adjusting for age, gender, race, ASA class, CCI score, estimated blood loss, baseline hemoglobin value, number of operated levels, and surgery type, logistic regression analysis revealed that patients with a nadir hemoglobin of 8-10 g/dL who were transfused had an independently higher risk of perioperative morbidity (odds ratio [OR] = 2.12; 95% confidence interval [CI], 1.24-3.64; P=0.006). Estimated additional costs associated with liberal trigger use, defined as a transfusion occurring in patients with a whole hospital stay nadir Hb of 8-10 g/dL, ranged from $202,675 to $700,151 annually. CONCLUSIONS: Transfusion using a liberal trigger is associated with increased morbidity, even after controlling for possible confounders. Our results suggest that modification of transfusion practice may be a potential area for improving patient outcomes and reducing costs.