Intraobserver and Interobserver Measurement Variability of the Rib-Vertebral Angle Difference.

Abstract

The rib-vertebral angle (RVAD) differentiates between progressive and resolving infantile idiopathic scoliosis (IIS) curves. Those with a RVAD<20 degrees often resolve without treatment, whereas those >20 degrees often progress and require treatment. The mathematical magnitude of RVAD measurement variability has not been described, and was thus the purpose of this study. It is important to know the reliability of RVAD measurements so as to understand changes that can be attributed to observer error alone. The purpose of the study was to mathematically determine the intraobserver and interobserver measurement variability of the RVAD difference in IIS. A convenience sample of 50 patients with IIS was selected. The RVAD was measured twice by 7 different observers separated by a minimum of 3 weeks. The 7 different readers chosen to measure the RVAD were 2 orthopaedic residents, 1 pediatric radiologist, and 4 attending pediatric orthopaedic surgeons. The same goniometer was used for all measurements. Intraobserver and interobserver measurement variability was determined using 3 well known used statistical methods. The 3 different methods used to determine intraobserver and interobserver variability using 95% prediction/confidence limits gave very similar results. Intraobserver variability was ∼24 degrees and the interobserver variability was ∼23 degrees. The utility of the RVAD in medical decision-making and counseling of patients with IIS should be approached with caution when used in isolation due to the large measurement variability. Level 3.