81 - Cobb Angle Measurements in Patients with Spinal Deformity: OFFICEPACS-...

#81 Cobb Angle Measurements in Patients with Spinal Deformity: OFFICEPACS-based Digital Method Compared with Traditional Manual Measurements

Epidemiology-Natural History

Poster Presented by: D. Ohnmeiss

Author(s):

I.H. Lieberman (1)
J. Sparks (1)
D.D. Ohnmeiss (2)
S. Verma-Kurvari (2)
R.D. Guyer (1)
J.E. Zigler (1)

(1) Texas Back Institute, Plano, TX, USA
(2) Texas Back Institute Research Foundation, Plano, TX, USA

Abstract

Introduction: As use of electronic imaging systems such as OFFICEPACS (Merge Healthcare, Chicago, IL) for making and evaluating spinal radiographs increases, reliability of measurements from such systems should be investigated. A few studies have compared manual to digital radiographic measurements in scoliosis patients; however, these involved specialized software not generally available to clinicians. The purpose of this study was to investigate the repeatability of Cobb angle measurements in spinal deformity patients using manual and OFFICEPACS methods.

Methods: Twenty patients with spinal deformity were randomly selected. Images were deidentified and made available to evaluators in traditional film x-ray format (measured with goniometer) and digitally on the OFFICEPACS system (measured using the Cobb angle tool). Nine evaluators, 5 spine surgery fellows and 4 attendings, were blinded to each others' scores and their own first set of scores. Repeatability for each method, as well as correlation between methods, was evaluated using intra-class correlation coefficients (ICC). A single-sample t-test was used to determine if the mean absolute value of differences between sets of intra-evaluator measurements were significantly different from 0 (0 representing perfect agreement).

Results: Inter-evaluator ICC was high for both techniques at 0.963 for computerized and 0.836 for manual. ICC for digital vs. manual measurements ranged from 0.882 to 0.897 for individual evaluators. The mean absolute values of differences between digital and manual measurements for each evaluator were significantly greater than 0, ranging from 2.5 to 5.6 degrees. There was no systematic tendency to consistently score higher, or lower, with either method. It was noted that the digital measurements were performed in less time than required for the manual measurements.

Discussion: Inter-evaluator ICCs for OFFICEPACS and manual methods were high, with the digital value being slightly greater. Intra-evaluator ICC for the two methods was also high. However, differences between sets of scores were significantly greater than 0 for all evaluators, indicating lack of measuring exactly the same number of degrees. However, the mean difference was generally less than 5 degrees, which is often used as a clinically relevant value indicating no change in the deformity. This study supports the use of digital methods for measuring Cobb angles in patients with spinal deformity, with no loss of reliability of the measurements compared with measuring traditional film radiographs. Accessing digital images is becoming a more widespread option with the increasing use of electronic medical records incorporating OFFICEPACS for image management.