Faculty Bibliography

BACKGROUND CONTEXT: Adult Spinal Deformity (ASD) surgery is associated with a high cost. Previous studies have suggested that the actual direct hospital cost of ASD surgery is higher than the Medicare Allowable (MA) rate. MA rates are becoming the benchmark reimbursement target for hospital accounting systems. It is important to determine what factors can predict which ASD patients have actual costs below Medicare reimbursement. PURPOSE: Our goal is to (1) determine if we can develop a model to predict when index episode of care (iEOC) surgical costs of ASD surgeries are below the MA threshold and (2) to identify potentially modifiable variables. STUDY DESIGN/SETTING: Retrospective analysis of a prospective multi-center ASD database. PATIENT SAMPLE: Adult spinal deformity patients undergoing surgical correction enrolled into a prospective, multicenter surgical database from 2008-2014. OUTCOME MEASURES: Primary outcome measure is iEOC costs and patient and hospital-specific MA reimbursements, but other baseline HRQOL outcome measures were collected such as Oswestry Disability Index (ODI), SF-36 PCS, and SF-36 MCS.
METHOD(S): From a prospective, multicenter ASD surgical database, patients undergoing long instrumented fusions (>4 level) with cost data were identified. Index episode of care (iEOC) cost was calculated by utilizing actual direct hospital cost. MA rates were calculated using the year-appropriate CMS Inpatient Pricer Payment System Tool and were hospital specific. Demographic, baseline HRQOL, radiographic and surgical variables were analyzed. A predictive model was developed to identify variables that can predict iEOCRESULT(S): Administrative direct cost data was obtained from 4 of 11 centers with a total of 195 patients included in the model. A total of 109 (55%) patients had iEOC below the MA threshold. There was significant variation across the 4 centers in both the mean iEOC cost ($56,788 to $78,878,p<0.001) and reimbursement ($40,623 to $91,351, p<0.001) which was seen across deformity-specific DRGs (453,454,456,457). Academic centers were more likely to have iEOC costs below reimbursement (66.7% vs 8.9%, p<0.001). The model showed excellent fit and calibration with an AUC of 94.48% [95%CI 91.59-97.38]. Negative predictors included number of levels fused (OR 0.61, 95%CI [0.48-0.75], p <0.001) and DRG 457 (OR 0.016, 95%CI [0.0-0.20], p = 0.0016) while having a revision surgery (OR 3.61, [1.06-13.58], p=0.045) and at an academic center (OR 19.6 95%CI [8.81-50.7], p <0.001) were positive predictors. The marginal effect for additional levels fused is -4.38%, meaning for each additional level fused, the likelihood of the iEOCCONCLUSION(S): There is significant institutional variation in ASD cost and reimbursement with an increased likelihood of 56.8% getting reimbursed more than the cost of surgery (iEOCCopyright

BACKGROUND CONTEXT: In the past, averages from published studies have been used for performance benchmarking. However, in practice and in large studies, usually only a small subset of patients fall within the "average" and therefore the average may not provide a valid benchmark rate. One can overperform or underperform the average and still attain the predicted performance benchmark. Predictive modeling can be used to help set performance standards based on a large number of variables including patients' specific factors, procedure details, use of osteogenic products, site specific variations and many others. This study attempts to use predictive modeling for pseudarthrosis performance benchmarking as a proof of concept in ASD surgery. PURPOSE: To use predictive modeling for performance benchmarking in high-volume centers and demonstrate that it is more appropriate than simply using averages. STUDY DESIGN/SETTING: Retrospective review of a prospective, multicenter ASD database from 11 different sites PATIENT SAMPLE: ASD operative patients with age >=18, Coronal Cobb >= 20deg, SVA >=5cm, PT>=25deg, and/or thoracic kyphosis (TK) >= 60deg with min 2-year follow-up. Exclusion criteria: having a revision for any indication other than pseudo in order to reduce confounding of potential pseudo as a result of the revision surgery. OUTCOME MEASURES: HRQOL scores: Oswestry Disability Index (ODI), Short form-36 (SF36), Scoliosis Research Society (SRS22), back/leg pain numerical rating scale (NRS). Radiographic values: max coronal cobb angle, coronal C7 plumb line, pelvic tilt (PT), mismatch between pelvic incidence and lumbar lordosis (PI-LL), thoracic kyphosis (TK), C7 sagittal vertical axis (SVA). Posterior fusion was graded to determine if pseudo has occurred within 2 years postoperative. Demographic, frailty, surgical including BMP and complications data were also collected.
METHOD(S): A prior validated and published pseudo predictive model has been constructed from 336 ASD patients with an accuracy of 91.3% and an AUC of 0.94 using 21 out of 82 total variables listed above. This model was deployed with an updated set of patients to determine the predicted pseudo rate for each individual surgical site. Actual pseudo rates of the 11 contributing surgical sites were compared to the predicted values as a means to assess for performance benchmarking.
RESULT(S): A total of 403 patients were included (80.1% Female, avg age 57.9+/-14.9 years) from a total of 502 operative patients with 99 excluded for having a revision for indication other than pseudo. A total of 129 (32.0%) had pseudo by 2 years. The overall pseudo rates per year were the following: 2008-20.0%, 2009-37.7%, 2010-31.4%, 2011-29.3%, 2012-36.7%, 2013-32.0%, 2014-26.6%, 2015-38.1%. Six sites had actual rates above the overall rate of 32% with 5 sites below. However, the predicted rates varied according to each site and included rates above/below the overall rate. All of the actual rates were larger than the predicted rates except for 4 sites in which the actual and predicted rates were the same.
CONCLUSION(S): The pseudo rate varied per year and per site. Predictive modeling was able to provide a customized pseudoarthosis rate for each site considering multiple variables allowing for performance benchmarking instead of the average. Even though a site was above the overall average rate, they may be predicted to have a higher rate given the type of patients being treated or procedures done at that site. FDA DEVICE/DRUG STATUS: This abstract does not discuss or include any applicable devices or drugs.
Copyright

BACKGROUND CONTEXT: Adult Spinal Deformity (ASD) surgery significantly improves disability in patients who are functionally limited. Few studies have been done examining the threshold at which surgery is unlikely to provide functional benefit in ASD patients. PURPOSE: To identify a preoperative threshold at which patients who are more active than this threshold are unlikely to have functional improvement following ASD surgery, and may experience functional decline. STUDY DESIGN/SETTING: Retrospective review of a multicenter database. PATIENT SAMPLE: A total of 592/760 (78%) eligible ASD patients (>5 levels fused) were identified in a multicenter database with minimum 2-year follow-up. OUTCOME MEASURES: Our primary outcome was baseline-2 year postoperative change in Scoliosis Research Society 22r (SRS-22r) Activity domain.
METHOD(S): The SRS-22r was administered at preoperatively and at follow-up visits through 2-years. A baseline SRS-22r Activity score threshold was identified at which patients were more likely to decline than improve at 2-years postoperatively. Categorical variables were compared with chi-squared test. Significance was set at 0.05.
RESULT(S): Mean age of patients was 59+/-14 years, 80% females. Patients had 11.2+/-4.3 levels fused; follow up of 3.8+/-0.9 years. A baseline SRS-22r threshold of 3.8 was identified at which 44 (52%) patients above this threshold declined and 40 (48%) had improved activity at 2-year follow-up. In patients who declined, mean change was -0.6+/-0.43 points. Patients above this threshold were younger, had less severe deformity, and less back/leg pain (p<0.001). Most patients worse than this threshold reported SRS-22r Pain as the most severe domain (51%), while most patients better than this threshold reported the SRS-22r appearance as the worst domain (58%) (p<0.001). Patients above and below this threshold were equally likely to be satisfied with their surgery at 2 years (p=0.92).
CONCLUSION(S): Baseline SRS-22r Activity score can be used to risk-stratify ASD patients by likelihood of functional improvement postoperatively. Patients doing better than this threshold are just as likely to be satisfied as more disabled patients at 2 years, likely due to improvement in appearance, pain or mental health rather than activity level. In patients with SRS-22r activity above 3.8, the potential for minimal functional benefit should be included in the shared decision making process. FDA DEVICE/DRUG STATUS: This abstract does not discuss or include any applicable devices or drugs.
Copyright

BACKGROUND CONTEXT: Sexual function is an important contributor to quality of life. Adult spinal deformity (ASD) patients have been shown to have sexual limitations due to their deformity. PURPOSE: This investigation sought to assess sexual dysfunction due to lumbar stiffness before and after fusion surgery. STUDY DESIGN/SETTING: Retrospective analysis of a multi-center, prospectively-collected, consecutive cohort of ASD patients. PATIENT SAMPLE: Only patients with 2-year follow-up were included. In total, 365 patients were included in this study, comprising 76 males and 289 females. OUTCOME MEASURES: The primary outcome in this study was the Lumbar Spine Disability Index (LSDI) question 10: "Choose the statement that best describes the effect of low back stiffness on your ability to engage in sexual intercourse." METHODS: Differences in sexual function between baseline and 2-year follow-up were assessed with a Wilcoxon-Mann-Whitney test. Patient factors associated with poor baseline sexual function were evaluated with multiple linear regression. The association between sexual function and HRQOL at baseline and 2-year follow-up was evaluated with multiple linear regression, adjusting for all factors previously included in analyses of baseline sexual function.
RESULT(S): Baseline LSDI sexual function scores averaged 2.7 (SD 1.3), which improved to 2.3 (SD 1.2) at 2-year postsurgical follow-up (p = 0.0009). Predictors of poorer baseline sexual function included older age, BMI, and higher Charlson Comorbidity Index (p<0.05 for all comparisons). After adjusting for confounding factors, worse LSDI sexual function score was strongly associated with worse ODI, SRS total, and SF-36 PCS at both baseline and 2-year follow-up (p<0.001 for all comparisons).
CONCLUSION(S): This study found that sexual dysfunction due to lumbar stiffness is strongly related to HRQOL measures such as ODI and SRS-22r total score. Further, sexual function was not diminished postoperatively, possibly due to reduced pain that accompanied improved stability. FDA DEVICE/DRUG STATUS: This abstract does not discuss or include any applicable devices or drugs.
Copyright

BACKGROUND CONTEXT: Existing health outcome (HRQL) metrics do not adequately capture disability from cervical deformity (CD) and do not correlate with cervical malalignment. In the novel Patient Generated Index (PGI) patients report their greatest difficulties related to their CD. These results were used to determine items that should be included in a CD-specific HRQL. PURPOSE: To utilize the PGI to reveal the aspects of CD disability not captured by existing HRQLs. STUDY DESIGN/SETTING: Retrospective review of a prospective CD database. PATIENT SAMPLE: A total of 45 CD patients. OUTCOME MEASURES: HRQL metrics: PGI, NDI, mJOA, EQ-5D.
METHOD(S): CD patients completed the PGI by describing aspects of their disability that bother them the most. The responses were weighted and scored. PGI responses were categorized into domains: Sagittal discomfort/range of motion (ROM), Activities of Daily Living (ADL), and Social Life/Hobbies. PGI scores and legacy HRQL metrics were correlated with alignment, pain, age, sex, BMI, and medical comorbidities. R2 values are reported for linear regression models that include the drivers significantly associated with each HRQL metric.
RESULT(S): Forty-five CD patients (mean cSVA: 51mm) including 12 PGI patients (mean cSVA: 62mm) were included for analysis. PGI scores were found to be driven significantly by age and C2 Slope (r2=0.50). NDI was driven significantly by neck pain, back pain, and BMI (r2=0.32). mJOA was driven significantly by Charlson Comorbidity Score, back pain and weight (r2=0.33). EQ5D was significantly driven by CBVA, age and T1 Slope (r2=0.78). When examining PGI domains, Sagittal Discomfort/ROM score was driven significantly by cSVA and age (r2=0.54). ADL score was driven by CBVA and a medical history of neuromuscular disease (r2=0.87). Social Life/Hobbies score was driven by Charlson Comorbidity Scores, a medical history of ankylosing spondylitis, and a medical history of connective tissue disease (r2=1.0). Horizontal Gaze/Walking Safety, Pain, and Neurologic Complaints did not correlate significantly with alignment, pain, demographic info or past medical history.
CONCLUSION(S): Legacy HRQLs do not adequately capture CD disability and do not correlate with cervical malalignment. In a CD cohort, PGI scores and EQ5D scores were driven significantly by sagittal alignment. However, mJOA and NDI were driven by pain and medical comorbidities. FDA DEVICE/DRUG STATUS: This abstract does not discuss or include any applicable devices or drugs.
Copyright

BACKGROUND CONTEXT: The Roussouly Classification system of sagittal spinal shape and the SRS-Schwab adult spinal deformity (ASD) classification system have become important indicators of spine deformity. No prior studies have examined the outcomes of matching both Roussouly type and improving in Schwab modifiers postoperatively. PURPOSE: Evaluate outcomes of matching Roussouly Type and improving in Schwab modifier following ASD surgery. STUDY DESIGN/SETTING: Retrospective review of single-center ASD database. PATIENT SAMPLE: A total of 103 ASD patients. OUTCOME MEASURES: Roussouly types, Schwab modifiers, Health Related Quality of Life scores(HRQLs): Minimal Clinical Important Difference for ODI, EQ5D, VAS Leg &Back Pain.
METHOD(S): Surgical ASD patients (SVA>=5cm, PT>=25degree, or TK >=60degree, >3 levels fused) >=18 years old with available baseline (BL) radiographic data at baseline (BL) and 1-year (1Y) were isolated in the single-center Comprehensive Spine Quality Database (Quality). Patients were grouped by two Roussouly types: (1)"theoretical" Roussouly type(Type 1: PI<45degree, LL apex below L4; Type 2: PI<45degree, LL apex above L4 L4-L5 space; Type 3: 45degree60degree); (2) "current" Roussouly type (1: SS<35degree, LL apex below L4; 2: PI<35degree, LL apex above L4-L5 space; 3: 35degree45degree), as previously published. One year (1Y) matched Roussouly: preoperative mismatched (Between 'actual' and 'theoretical' patients that matched at 1Y. Schwab modifiers at BL were identified: non-, moderate and severe deformity (0, +, ++) for PT, SVA, and PI-LL. Schwab improvement was defined as a decrease in a modifier at one year.
RESULT(S): A total of 103 ASD patients (61.8yrs, 63.1%F, 30kg/m2). By surgical approach, 79.6% posterior, 10.7% combined, 2.9% anterior). Average levels fused: 4.6. BL breakdown of 'current; Roussouly type: 28% Type 1, 25.3% Type 2, 32.0% Type 3, 14.7% Type 4. BL Roussouly mismatch: 65.3%. Breakdown BL Schwab modifiers: PT (0: 8.7%, +: 41.7%, ++: 49.5%), SVA (0: 29.7%, +: 20.3%, ++: 50%), PI-LL mismatch (0: 28.2%, +: 25.2%, ++: 46.6%). At one year, 19.2% of patients matched Roussouly target type, while according to Schwab modifiers, 12.6% improved in SVA, 42.7% in PI-LL, and 45.6% in PT. Patients who both met Roussouly type and improved in a Schwab by the modifiers: 9 PT (8.7%), 8 PI-LL (7.8%), 2 SVA (1.9%). There were 2 patients (1.9%) who met their Roussouly type and improved in all 3 Schwab modifiers. One year (1Y) matched Roussouly patients improved more in HRQLs (MCID for ODI, EQ5D, VAS Leg/Back Pain), when compared to mismatched Roussouly, but was not significant(P>0.05). Match Roussouly and improvement in PT Schwab met MCID for EQ5D significantly more (33.3% vs 10.6%, p=0.050). Matched Roussouly and PI-LL Schwab had more patients meet MCID for all HRQLs, yet none were significant, p>0.05. Matched Roussouly and improvement in SVA Schwab met MCID for ODI significantly more (p=0.024).
CONCLUSION(S): Patients who both matched Roussouly sagittal spinal type and improved in SRS-Schwab modifiers had superior patient-reported outcomes at 1-year. Utilizing both classification systems in surgical decision making can optimize postop patient outcomes. FDA DEVICE/DRUG STATUS: This abstract does not discuss or include any applicable devices or drugs.
Copyright

BACKGROUND CONTEXT: The current proposed cervical deformity (CD) classification system including cutoffs for C2-C7 sagittal vertical axis(cSVA), T1 slope minus C2-C7 lordosis (TS-CL), Chin Brow Vertical Angle (CBVA), and the modified Japanese Orthopaedic Association scale(mJOA) are based upon a modified Delphi approach and expert opinion. PURPOSE: To investigate the relationship between cervical parameters and the Health Related Quality of Life (HRQL) measure, mJOA. STUDY DESIGN/SETTING: Retrospective review of a prospective, consecutively enrolled CD database. PATIENT SAMPLE: A total of 123 CD patients. OUTCOME MEASURES: CD modifiers; HRQL Instrument: mJOA; Radiographic thresholds for moderate (M) and severe (S) deformity.
METHOD(S): Included: surgical adult CD patients (C2-C7 Cobb>10degree, CL>10degree, cSVA>4cm, or CBVA>25degree). Radiographic parameters measured: C2-C7 lordosis (CL), T1 slope, C2-C7 SVA, TS-CL, C2-T3 angle, C2 slope, pelvic tilt (PT). Measured HRQLs: Reported mJOA scores, categorized by Tetreault et al.: 18 as none, 15-17 as mild, 12-14 as moderate and <12 as severe. Other modifiers assessed: McGregor slope (MGS), CBVA and frailty. Statistical analyses were performed to determine correlations between HRQLs and possible modifiers. After determining data followed a parametric distribution using the Shapiro Wilk Normality test (p=0.15, p>0.05), Pearson correlations were run for all combinations. For significant correlations, linear and binary logistic regressions were performed to determine a possible threshold of radiographic measures for which the correlation with mJOA scores was most significant. A mJOA score of 14 and <12 that are reported cut off values for moderate and severe disability.
RESULT(S): A total of 123 CD patients were included (60.5+/-10.1 years, 65% female, 29.1+/-8.2 kg/m2). The average Charlson Comorbidity Index (CCI) score was 0.90. Cohort surgical factors included, by approach, 17.1% anterior, 50.4% posterior and 32.5% combined. Average anterior levels fused was 3.36, while the average posterior levels fused was 8.93, and the total average was 8.05 levels fused. The mean total operative time was 534.3 minutes, with an estimated blood loss (EBL) of 890.6 ccs. Baseline mJOA score was 13.54+/-2.8. Pearson correlations determined significant correlation between baseline mJOA scores and McGregor's Slope (-0.236, p=0.015), TS-CL (-0.246, p=0.006), CL (0.225, p=0.012), C2-T3 (0.180, p=0.046), C2 Slope (-0.234, p=0.009), and frailty (0.517, p<0.001), and no significant correlation with T1 slope, cSVA, CBVA and PT. Logistic regression models assigned values as binary variables greater or less than a predicted threshold value, tested at increments. For significant baseline factors from the Pearson correlation, the following rough thresholds were predicted: MGS (M: -12 to -9degree and 0 to 19degree, p=0.020; S: >19degree and <-12degree, x2= 4.291, p=0.036), TS-CL (M: 26degreeto 45degree, p=0.307; S: >45degree, x2= 7.8, p=0.005), CL (M: -21degree to 3degree, x2= 8.947, p=0.004; S: >3degree, x2 =9.3, p=0.009), C2-T3 (M: -35degree to -25degree, x2 = 5.485, p=0.046; S: >-25degree, x2 = 4.1, p=0.041), C2 Slope (M:33degreeto 49degree, p=0.122; S: >49degree, x2=5.7, p=0.008), and Frailty (Mild: 0.18 to 0.27, p=0.129; Severe: >0.27, p=0.002).
CONCLUSION(S): Novel thresholds were established for McGregor's Slope, TS-CL, C2-T3 angle, C2 Slope and frailty. Each correlated with moderate or severe neurologic myelopathy HRQL by way of mJOA score. These cut-off values can be utilized in classifying cervical deformity. FDA DEVICE/DRUG STATUS: This abstract does not discuss or include any applicable devices or drugs.
Copyright

BACKGROUND CONTEXT: In the current value-based health care climate where spinal surgery is shifting to the ambulatory setting, factors influencing postoperative patient length of stay (LOS) have significance to both surgeons and hospital administrators. Underlying patient factors including the diagnosis of radiculopathy and myelopathy have not been investigated in this context. PURPOSE: Identify predictors extended LOS(E-LOS) between myelopathy(M), radiculopathy(R), and patients with both (MR). STUDY DESIGN/SETTING: Retrospective review of a single-center stereographic database. PATIENT SAMPLE: A total of 718 surgical cervical spine patients. OUTCOME MEASURES: Postoperative LOS, patient factors, preoperative HRQL, complications, predictors of E-LOS.
METHOD(S): Surgical cervical spine patients >=18yrs diagnosed with M or R primary diagnoses were isolated in the single-center Comprehensive Spine Quality Database (Quality). Patients were stratified by surgical approach: Anterior, Posterior or Combined. Top-quartile LOS values were labeled as extended. M and R patients were compared using chi-squared and independent samples t-tests, as appropriate. Univariate tests assessed differences in baseline patient-related and surgical data between M and R, and extended/non-extended LOS. Univariate/multivariate analyses were run to assess predictive factors of E-LOS in each diagnosis group. Regression with stepwise model selection was employed to explore factors potentially significant in predicting LOS.
RESULT(S): A total of 718 patients (54.5 years, 41.1%F, 29.1kg/m2). Mean CCI score: 1.11. Within the cohort, 177 patients (24.7%) had a diagnosis of myelopathy, 383 (53.3%) radiculopathy, and 22% with a diagnosis of myeloradiculopathy. Patients with M primary diagnosis were significantly older (62.2 vs 49.8yrs, p<0.001) and had a greater CCI score (1.64 vs 0.82, p<0.001) when compared to R patients. By approach: 76.7% anterior (57.6% of M, 90.6%R, 64.6%MR; p<0.001), 16.4% (35%M, 6%R, 20.9%MR; p<0.001) posterior, 6.5% (6.8%M, 3.4%R, 13.9%MR; p<0.001) combined. Average LOS: M(3.8days), R(1.5 days), MR(2.9 days) p<0.001. LOS for anterior approach in each diagnosis was as follows, M: 2.21, R: 1.21, MR: 1.69 days, p<0.001. Meanwhile, posterior approach LOS, M:6.06, R:2.91, MR:5.0, p<0.001; combined approach M: 5.17, R: 6.23, MR: 5.59, P=0.881. A total of 195 patients were categorized as E-LOS (Avg: 5.87 days), 87 M, 43 R, 65 MR. Major surgical approach of E-LOS for M (60.9%) and MR (44.6%) was posterior; whereas R E-LOS patients majorly underwent anterior procedures (53.5%). Generalized linear regression modeling found that the following combination of factors predicted E-LOS in R patients (R2=0.736, p=0.003):BMI, durotomy, CCI, anterior and combined approaches, and cardiac complications. An additional model discovered the predictors of E-LOS in M patients (R2= 0.312, p<0.001): age, hypertension, CCI, anterior and combined approaches, intraoperative complications, neuro complications, ileus, and return to OR in 90 days. Lastly, the model for E-LOS in MR patients consisted of (R2 = 0.267, p=0.001): age, durotomy, BL EQ5D, hypertension, posterior and combined approaches and postoperative complications, specifically neuro.
CONCLUSION(S): Independent of surgical approach, patients with a primary diagnosis of myelopathy, though older aged and higher comorbidity profile, had consistently longer overall postop LOS when compared to radiculopathy or myeloradiculopathy patients. The heightened risk in myelopathy patients for extended LOS should be considered when determining admission status for patients undergoing cervical spine surgery. FDA DEVICE/DRUG STATUS: This abstract does not discuss or include any applicable devices or drugs.
Copyright

BACKGROUND CONTEXT: Distal junctional kyphosis (DJK) is a major concern following cervical deformity (CD) correction, leading to failed realignment and revision surgery. Undercorrection may be the major factor associated with DJK, however it is difficult to assess because the DJK erodes the correction and worsens traditional measures like cSVA. In-construct measures have been proposed that measure alignment within the fusion separate from the DJK and subjacent compensation. PURPOSE: To simulate cervical deformity corrections using surgical planning software to determine if undercorrection results in DJK. STUDY DESIGN/SETTING: Retrospective analysis of a prospective, multicenter CD database. PATIENT SAMPLE: A total of 69 CD patients. OUTCOME MEASURES: Radiographic measures: C2 plumblines, in-construct measures and DJK angle change.
METHOD(S): A prospective database of operative CD patients was analyzed for severe DJK (kyphosis change>20degree in LIV to LIV-2) and traditional DJK (change>10degree). C2LIV-Tilt (angle of a line from the centroid of C2 to the centroid of the lower instrumented vertebra and a line along the posterior vertebral body wall of the LIV) was measured postoperatively and the correction was simulated in the preoperative X-ray to match the C2LIV-Tilt using planning software. Linear regression analysis using C2 pelvic angle (CPA) and Pelvic Tilt (PT) determined the simulated PT to match the virtual CPA. SVA measures were compared in patients with severe and traditional DJK and no DJK. Linear regression analysis was used to determine the C2-T4 and C2-T10 Tilts that correspond to DJK=10degree and cSVA=4cm.
RESULT(S): A total of 69 CD patients (mean age 61, 60% female) were included. Severe and traditional DJK occurred in 11 (16%) and 22 (32%) patients; 3 (4%) required DJK revision. Simulated corrections demonstrated that severe and traditional DJK pts had worse alignments compared to no DJK pts: cSVA (42.5 vs 33.0 vs 23.4mm, p<.001), C2-T3 SVA (74.5 vs 61.8 vs 41.4mm, p<.001), C2-LIV SVA (68.9 vs 57.3 vs 36.8mm, p<.001). Linear regression revealed a predictive relationship between in-construct measures (C2T4-Tilt and C2T10-Tilt) and cSVA and change in DJK (all R>.57, p<.001). cSVA=4cm corresponded to C2T4-Tilt of 10.4degree and C2T10-Tilt of 28.0degree. DJK change=10degree corresponded to C2T4-Tilt of 5.8degree and C2T10-Tilt of 20.1degree. Severe DJK pts had the worst postoperative alignment by all measures including cSVA, TSCL, CPA, C2LIV-Tilt, (all p<.001).
CONCLUSION(S): Simulated cervical deformity corrections demonstrated that severe DJK patients have insufficient corrections compared to patients without DJK. In-construct measures assess sagittal alignment within the fusion separate from DJK and subjacent compensation. They can be useful as intraoperative tools to gauge the adequacy of CD correction. FDA DEVICE/DRUG STATUS: This abstract does not discuss or include any applicable devices or drugs.
Copyright

BACKGROUND CONTEXT: Severe cervical deformity (CD) can be a source of severe disability. Surgical correction of the deformity can markedly improve a patient's quality of life and functional status. However, high rates of complications have been reported, one of the most serious being distal junctional kyphosis (DJK). Although researchers have identified risk factors for DJK, no model has been proposed that predicts the magnitude of DJK. PURPOSE: To develop a model that can accurately predict the DJ angle following CD correction using radiographic measurements that can be obtained pre- and intraoperatively. STUDY DESIGN/SETTING: Retrospective review of a prospectively-collected CD database. PATIENT SAMPLE: A total of 131 preoperative patients with cervical deformity. OUTCOME MEASURES: DJK angle, reoperation for DJK METHODS: A prospective database of operative CD patients (C2-C7 Cobb>10degree, CL>10degree, cSVA>4cm or CBVA>25degree) with clinical and radiographic data at preoperative and any postoperative interval (3m, 6m, 1y, 2y). Patients with a history of fusion below L4 were excluded. DJK angle was defined as the Cobb from LIV to LIV-2, traditional DJK (tDJK) was a DJK angle change >10degree; a change >20degree was defined as severe DJK (sDJK) was a DJK angle change >20degree.Models were developed using a randomly selected 66.6% of patients. Pre- and intraoperative radiographic parameters significantly associated with DJK were identified and ranked in order of importance via a conditional variable importance table that used a nonreplacement sampling set of 4,000 conditional inference trees. Linear regression models were developed using the factors most strongly associated with postoperative DJK angle. The models were then validated in the remaining 33.3% of patients.
RESULT(S): A total of 131 patients were included (60 +/-10y, 62% female). Mean follow-up was 14+/-8 months. The most common LIV was T2 (20.6%) with an LIV below T2 in 44%. The mean postoperative DJK angle was 14.6degree +/-14degree. 35% of patients developed DJK (11% sDJK). Five patients underwent a subsequent revision surgery due to DJK. The model development subset included 84 patients and did not differ significantly from the validation subset in terms of radiographic or clinical parameters. The variables identified as being most significant were: Intraoperative change of CL (DELTACL) and C2-LIV Tilt (DELTAC2-LIV), and preoperative DJK angle (DJK-Apre). The initial model including DELTACL and DELTAC2-LIV accounted for 19.6% of the variability in postoperative DJKA (r=.470, p<0.001). The following model, which included preoperative DJKA, accounted for 37.0% of the variability (R=.609, p<0.001).DJKA = 9.365 + (.123* DELTACL) - (0.315* DELTAC2-LIV) - (0.504*DJKApre) This equation was then validated in the remaining 47 patients. It shows that as you improve alignment by increasing CL and decreasing the C2-LIV tilt, the DJK angle increases. Predicted and actual postoperative DJKA values were highly correlated (R=0.871, R2= 0.759, p<0.001). The model RMSE was 0.97. The model had better specificity for sDJK vs tDJK (94% vs 57%), accuracy (85% vs 61%), and NPV (89% vs 72%). But worse sensitivity (17% vs 67%) and PPV (29% vs 50%).
CONCLUSION(S): DJK is a serious, but fairly common complication following correction of CD and can lead to pain and disability necessitating revision surgery. Through the use of pre- and intraoperative measurements, surgeons can reliably predict the magnitude of DJK following correction. In order of importance, the variables that increased the DJK angle were, preoperative DJK angle, the magnitude of the correction within the construct and the change in CL. The models developed in this study, which reliably predict postoperative DJK angle and severe DJK, can be easily applied in a clinical setting when planning postoperative alignment for CD correction. FDA DEVICE/DRUG STATUS: This abstract does not discuss or include any applicable devices or drugs.
Copyright