July 2019
Volume 60, Issue 9
Open Access
ARVO Annual Meeting Abstract  |   July 2019
Predicting Risk of Perioperative Ischemic Optic Neuropathy: A Study Using the US National Inpatient Sample
Author Affiliations & Notes
  • Steven Roth
    Anesthesiology, Univ of Illinois, Chicago, Illinois, United States
    Ophthalmology and Visual Science, University of Illinois, Chicago, Illinois, United States
  • Shikhar Shah
    Anesthesiology, Walter Reed National Military Medical Center, Bethesda, Maryland, United States
  • Yi-Fen Chen
    Clinical and Translational Sciences, University of Illinois at Chicago, Chicago, Illinois, United States
  • Heather Moss
    Ophthalmology and Visual Science; Neurology and Neurological Sciences, Stanford University, Palo Alto, California, United States
  • Daniel Rubin
    Anesthesia and Critical Care, University of Chicago, Chicago, Illinois, United States
  • Charlotte Joslin
    Ophthalmology and Visual Science, University of Illinois, Chicago, Illinois, United States
  • Footnotes
    Commercial Relationships   Steven Roth, Dr Roth has served as an expert witness in cases of perioperative visual loss (C); Shikhar Shah, None; Yi-Fen Chen, None; Heather Moss, None; Daniel Rubin, None; Charlotte Joslin, None
  • Footnotes
    Support  NIH Grants EY02477 and EY02477-01S1 to SR
Investigative Ophthalmology & Visual Science July 2019, Vol.60, 4805. doi:
  • Views
  • Share
  • Tools
    • Alerts
      ×
      This feature is available to authenticated users only.
      Sign In or Create an Account ×
    • Get Citation

      Steven Roth, Shikhar Shah, Yi-Fen Chen, Heather Moss, Daniel Rubin, Charlotte Joslin; Predicting Risk of Perioperative Ischemic Optic Neuropathy: A Study Using the US National Inpatient Sample. Invest. Ophthalmol. Vis. Sci. 2019;60(9):4805.

      Download citation file:


      © ARVO (1962-2015); The Authors (2016-present)

      ×
  • Supplements
Abstract

Purpose : Perioperative schemic optic neuropathy (ION) in spinal fusion surgery causes severe vision loss, with no effective treatment. We sought to develop a validated predictive model. We also tested hypotheses that incidence of perioperative ION in spinal fusion is decreasing and that certain pre-existent eye conditions increased risk.

Methods : Discharge data from NIS (National Inpatient Sample) were utilized. Our Institutional Review Board deemed this study “exempt." ICD-9CM diagnostic and procedure codes were used. In the NIS for 1998-2014, posterior thoracic, lumbar and sacral spine fusion were identified. Univariate and multivariable logistic regression identified risk factors in the weighted NIS sample. By splitting our sample, into a training (67% of all spine fusion cases) and testing (33%) dataset, the "split sample" method, we developed a predictive model and internally validated it. Logistic regression created a predictive model from the training data set using backwards model selection and p < 0.05. Once this most parsimonious model was defined, we developed a scoring scheme using the β-coefficients (or log of odds ratios), which had been rounded to the nearest integer. The sum of this predictive score was then evaluated in the test data for sensitivity, specificity, and area under the receiver operator (ROC) characteristic curve (AUC).

Results : ION was estimated at 0.95/10,000 spinal fusion discharges. During the study period of 1998 to 2014, the incidence of ION in spine fusion significantly decreased over time (0.83 per 2 y, 95% confidence intervals (CI):0.74-0.94, p = 0.002). Factors significantly associated with ION were more vertebral levels in surgery (2-3 as reference; OR 2.92 for 4-8, CI:1.48-5.78, p = 0.002; OR 4.50 for ≥ 9, CI:1.40-14.54, p = 0.012), glaucoma (OR 5.42, CI:1.80-16.30, p = 0.003), and operative bleeding (OR 3.92, CI:1.10-13.98, p = 0.035). ROC with AUC = 0.68 and score = 0 had 100% sensitivity to predict that a spinal fusion discharge was not associated with perioperative ION (Figure 1).

Conclusions : The predictive model had excellent sensitivity and could enable screening for patients at higher risk of ION and to provide more personalized informed consent. New risk factors identified in this study include more vertebral levels, glaucoma, and operative bleeding.

This abstract was presented at the 2019 ARVO Annual Meeting, held in Vancouver, Canada, April 28 - May 2, 2019.

 

Predictive model for perioperative ION in spine surgery, AUC = 0.68

Predictive model for perioperative ION in spine surgery, AUC = 0.68

×
×

This PDF is available to Subscribers Only

Sign in or purchase a subscription to access this content. ×

You must be signed into an individual account to use this feature.

×