Predicting the combined occurrence of poor clinical and radiographic outcomes following cervical deformity corrective surgery

Abstract

<jats:sec><jats:title>OBJECTIVE</jats:title><jats:p>Cervical deformity (CD) correction is clinically challenging. There is a high risk of developing complications with these highly complex procedures. The aim of this study was to use baseline demographic, clinical, and surgical factors to predict a poor outcome following CD surgery.</jats:p></jats:sec><jats:sec><jats:title>METHODS</jats:title><jats:p>The authors performed a retrospective review of a multicenter prospective CD database. CD was defined as at least one of the following: cervical kyphosis (C2–7 Cobb angle &gt; 10°), cervical scoliosis (coronal Cobb angle &gt; 10°), C2–7 sagittal vertical axis (cSVA) &gt; 4 cm, or chin-brow vertical angle (CBVA) &gt; 25°. Patients were categorized based on having an overall poor outcome or not. Health-related quality of life measures consisted of Neck Disability Index (NDI), EQ-5D, and modified Japanese Orthopaedic Association (mJOA) scale scores. A poor outcome was defined as having all 3 of the following categories met: 1) radiographic poor outcome: deterioration or severe radiographic malalignment 1 year postoperatively for cSVA or T1 slope–cervical lordosis mismatch (TS-CL); 2) clinical poor outcome: failing to meet the minimum clinically important difference (MCID) for NDI or having a severe mJOA Ames modifier; and 3) complications/reoperation poor outcome: major complication, death, or reoperation for a complication other than infection. Univariate logistic regression followed by multivariate regression models was performed, and internal validation was performed by calculating the area under the curve (AUC).</jats:p></jats:sec><jats:sec><jats:title>RESULTS</jats:title><jats:p>In total, 89 patients with CD were included (mean age 61.9 years, female sex 65.2%, BMI 29.2 kg/m<jats:sup>2</jats:sup>). By 1 year postoperatively, 18 (20.2%) patients were characterized as having an overall poor outcome. For radiographic poor outcomes, patients’ conditions either deteriorated or remained severe for TS-CL (73% of patients), cSVA (8%), horizontal gaze (34%), and global SVA (28%). For clinical poor outcomes, 80% and 60% of patients did not reach MCID for EQ-5D and NDI, respectively, and 24% of patients had severe symptoms (mJOA score 0–11). For the complications/reoperation poor outcome, 28 patients experienced a major complication, 11 underwent a reoperation, and 1 had a complication-related death. Of patients with a poor clinical outcome, 75% had a poor radiographic outcome; 35% of poor radiographic and 37% of poor clinical outcome patients had a major complication. A poor outcome was predicted by the following combination of factors: osteoporosis, baseline neurological status, use of a transition rod, number of posterior decompressions, baseline pelvic tilt, T2–12 kyphosis, TS-CL, C2–T3 SVA, C2–T1 pelvic angle (C2 slope), global SVA, and number of levels in maximum thoracic kyphosis. The final model predicting a poor outcome (AUC 86%) included the following: osteoporosis (OR 5.9, 95% CI 0.9–39), worse baseline neurological status (OR 11.4, 95% CI 1.8–70.8), baseline pelvic tilt &gt; 20° (OR 0.92, 95% CI 0.85–0.98), &gt; 9 levels in maximum thoracic kyphosis (OR 2.01, 95% CI 1.1–4.1), preoperative C2–T3 SVA &gt; 5.4 cm (OR 1.01, 95% CI 0.9–1.1), and global SVA &gt; 4 cm (OR 3.2, 95% CI 0.09–10.3).</jats:p></jats:sec><jats:sec><jats:title>CONCLUSIONS</jats:title><jats:p>Of all CD patients in this study, 20.2% had a poor overall outcome, defined by deterioration in radiographic and clinical outcomes, and a major complication. Additionally, 75% of patients with a poor clinical outcome also had a poor radiographic outcome. A poor overall outcome was most strongly predicted by severe baseline neurological deficit, global SVA &gt; 4 cm, and including more of the thoracic maximal kyphosis in the construct.</jats:p></jats:sec>

Association of Spinal Alignment Correction With Patient-Reported Outcomes in Adult Cervical Deformity: Review of the Literature

“…We identified 19 studies that studied a correlation between cervical parameters and PROs used to assess pain, disability, and quality of life in adult cervical deformity surgery (Figure 1). Across the 19 studies, 10 studies were conducted by the International Spine Study Group (ISSG) representing 19 sites in the U.S, 2 sites in Japan, and 1 site in Canada 3,[14][15][16][17][18][19][20][21] . The remaining studies were conducted in single institutions in the U.S 4,22-24 , Lebanon 25 , China 25 , Japan 26 , and South Korea 27,28 .…”

“…Cervical C2-C7 sagittal vertical axis (C2-C7 SVA) was investigated in 9 of the included studies 4,16,19,21,24,[26][27][28][29] , including 3 studies by the ISSG 16,19,30 . Meta-analyses of correlation coefficients showed that C2-C7 SVA and NDI correlated fairly [Random Effect; Zr = 0.31 [95% CI: -0.03, 0.58]; P < .001) 16,24,27,28,31 .…”

“…Eight studies investigated the relationship between cervical lordosis minus T1 slope (TS-CL) and PROs 4,16,19,22,24,25,27,28 . Meta-analyses of correlation coefficients showed that NDI and EQ-5D correlated poorly with TS-CL [NDI Random Effect; Zr = 0.04 [95% CI: -0.23, 0.30]; P =.001] and EQ-5D Random Effect; r = 0.07 [95% CI: -0.24, 0.37]; P =0.02) (Figure 3 25 at baseline.…”

Narrative Review of Predictive Analytics of Patient-Reported Outcomes in Adult Spinal Deformity Surgery

“…Predictors of poor NDI scores included the presence of osteoporosis, a worse baseline functional status, baseline pelvic tilt (PT) >20°, greater than 9 levels of thoracic kyphosis, and an elevated C2-T3 SVA or global SVA >4cm. 19 Bortz et al used conditional inference tree modeling to predict nonroutine discharge following cervical deformity surgery, defined as discharge to a rehabilitation centers rather than home. Their model found that patients who underwent nonroutine discharge had poorer overall health, as measured by the EuroQol-five dimensions (EQ-5D).…”

“…[8][9][10][11] Within ASD surgery, predictive analytics has also been used to predict the need for blood transfusion, hospital length of stay, complications, pseudoarthrosis, and catastrophic costs, [12][13][14][15][16] as well as postoperative PROs. [17][18][19][20][21][22][23] Given the numerous variables that contribute to outcomes and patient satisfaction with ASD, predictive analytics provides a valuable tool to analyze large data sets with unclear links of variables to outcomes. As a result, there has been substantial heterogeneity in studies on this topic, which suggests the importance of a review summarizing the existing evidence.…”