Investigator analytic repeatability of two new intervertebral motion biomarkers for chronic, nonspecific low back pain in a cohort of healthy controls

Abstract

Background: Understanding the mechanisms underlying chronic, nonspecific low back pain (CNSLBP) is essential to advance personalized care and identify the most appropriate intervention. Recently, two intervertebral motion biomarkers termed "Motion Sharing Inequality" (MSI) and "Motion Sharing Variability" (MSV) have been identified for CNSLBP using quantitative fluoroscopy (QF). The aim of this study was to conduct intra- and inter-investigator analytic repeatability studies to determine the extent to which investigator error affects their measurement in clinical studies.

Methods: A cross-sectional cohort study was conducted using the image sequences of 30 healthy controls who received QF screening during passive recumbent flexion motion. Two independent investigators analysed the image sequences for MSI and MSV from October to November 2018. Intra and inter- investigator repeatability studies were performed using intraclass correlations (ICC), standard errors of measurement (SEM) and minimal differences (MD).

Results: Intra-investigator ICCs were 0.90 (0.81,0.95) (SEM 0.029) and 0.78 (0.59,0.89) (SEM 0.020) for MSI and MSV, respectively. Inter-investigator ICCs 0.93 (0.86,0.97) (SEM 0.024) and 0.55 (0.24,0.75) (SEM 0.024). SEMs for MSI and MSV were approximately 10 and 30% of their group means respectively. The MDs for MSI for intra- and inter-investigator repeatability were 0.079 and 0.067, respectively and for MSV 0.055 and 0.067.

Conclusions: MSI demonstrated substantial intra- and inter-investigator repeatability, suggesting that investigator input has a minimal influence on its measurement. MSV demonstrated moderate intra-investigator reliability and fair inter-investigator repeatability. Confirmation in patients with CNSLBP is now required.

Keywords: Back pain; Biomarkers; Fluoroscopy; Kinematics; Repeatability.

Fig. 1

Apparatus for passive recumbent lumbar spine quantitative fluoroscopy image acquisition

Fig. 2

Reference templates (yellow) and tracking templates (green) were created on the first image of each sequence to allow for automated frame-to-frame tracking of the vertebral bodies in subsequent images of the sequence

Fig. 3

Derivation of motion sharing inequality (MSI) and motion sharing variability (MSV) from a representative QF image sequence obtained from one participant during lumbar flexion and return. Absolute intervertebral rotations, where the forward flexion direction is considered a decrease in intervertebral angle (a), are transformed into proportional intervertebral rotations (b), which allow for the calculation of the ranges of the proportional intervertebral movement. MSI is the average of the range of proportional intervertebral movement, while MSV is the standard deviation of the range of proportional intervertebral movement (c)

Fig. 4

Scatterplots and intraclass correlation coefficients (ICCs) for a intra-investigator repeatability for motion sharing inequality (MSI), b inter-investigator repeatability for MSI, c intra-investigator repeatability for motion sharing variability (MSV), and d inter-investigator repeatability for MSV with standard errors of measurement (SEMs) and minimal differences (MDs). The dashed line represents the line of identity between observations (a and c) or investigators (b and d)