340 - Predictive Model for Patient-reported Outcomes Scores Following Cervic...

Oral Posters: Adult Spinal Deformity

Presented by: D. Montes - View Audio/Video Presentation (Members Only)


P. Passias(1), C. Oh(2), C. Jalai(1), S. Horn(1), G. Poorman(1), A. Sure(1), J. Tishelman(1), R. Lafage(3), B. Diebo(4), J. Scheer(5), J. Smith(6), C. Shaffrey(6), T. Protopsaltis(1), H.J. Kim(3), R. Hart(7), V. Lafage(3), C. Ames(8), D. Montes(1), International Spine Study Group

(1) NYU Hospital for Joint Diseases, Orthopaedic Surgery, New York, NY, United States
(2) NYU Langone Medical Center, New York, NY, United States
(3) Hospital for Special Surgery, New York, NY, United States
(4) SUNY Downstate, Orthopaedic Surgery, Brooklyn, NY, United States
(5) University of Illinois at Chicago, Neurosurgery, Chicago, IL, United States
(6) University of Virginia Medical Center, Neurosurgery, Charlottesville, VA, United States
(7) Oregon Health and Science University, Orthopaedic Surgery, Portland, OR, United States
(8) University of California at San Francisco, Neurological Surgery, San Francisco, CA, United States


Background: Predicting and optimizing patient-reported outcomes following surgery for cervical deformity correction, wherein myelopathy is a chief diagnosis, is particularly important. However, most predictive models have focused on ideal post-operative alignment parameters, without consideration for patient-reported outcomes.

Purpose: This study incorporates baseline demographic factors, pre-operative alignment, and surgical variables into a predictive model for acute post-operative improvements in health-related quality of life (HRQoL).

Study Design/Setting: Retrospective review of a prospective multi-center database.

Patient Sample: 84 surgical patients with cervical spine deformity (CSD) were identified (mean age: 63.2yrs, 61.9% F).

Outcome Measures: 3 month (M) mJOA (Modified Japanese Orthopaedic Association) scores were considered as raw values. Patients meeting 3M mJOA MCID based on published values were also evaluated.

Methods: Inclusion criteria were CSD patients ≥18yrs with pre- and 3M post-operative radiographic and health-related quality of life (HRQL) for the mJOA. Univariate logistic regression assessed associations of independent variables with outcome values. Subsequent multivariate regression models with backward stepwise selection were performed to generate a dataset-specific prediction model. Final prediction model for linear scores (3M mJOA) was selected based on lowest Akaike information criterion (AIC) value. Dichotomous outcomes (percent of patients meeting 3M mJOA MCID) prediction was internally validated by calculating the area under the curve (AUC), reported with the 95% CI.

Results: 84 surgical CSD patients were included. The final model for predicting increased 3M mJOA raw scores yielded an AIC = 51.2. The following demographic and operative factors were included in the final model: pre-op mJOA score, BMP-2 use, posterior osteotomy at T1, a prior history of cervical fusion, and baseline thoracic kyphosis (p< 0.03 all cases). At 3 months post-operative, there were 13 patients (15.5%) that attained MCID for the mJOA. The likelihood of reaching 3M mJOA MCID was also accurately predicted (AUC=80.5% [94.13-66.77]), based on the following variables: lower LIV and increased baseline mJOA score (p< 0.04).

Conclusions: Acute post-operative improvements in myelopathy scores among CSD patients can be predicted with high accuracy using a combination of operative and radiographic parameters. Proper assessment of such factors should be taken at pre-op consultation to optimize patient outcomes. Levels of Evidence: III