Measurement of lumbar lordosis. Evaluation of intraobserver, interobserver, and technique variability.

Abstract

Sixty radiographs were measured on two separate occasions by three physicians using four different techniques to evaluate the reliability and reproducibility of the measurement of lumbar lordosis. To evaluate clinical methods of measuring lumbar lordosis, determining intraobserver and interobserver reliability. Several different methods are used to measure lumbar lordosis. The reliability and reproducibility of these has not been well studied. Sixty lateral full spine radiographs were obtained, labeled, and the lumbar lordosis measured independently by three practitioners who routinely perform these measurements. Four measurement techniques were used. These included measurements from the inferior endplate of T12 to the superior endplate of S1; the superior endplate of L1 to the superior endplate of S1; the inferior endplate of T12 to the inferior endplate of L5; and the superior endplate of L1 to the inferior endplate of L5. The measurements then were repeated after relabeling. Intraobserver reliability coefficients ranged from 0.83 to 0.92, indicating excellent reproducibility. Ninety-two percent of repeat measures were within 10 degrees. High overall and pairwise agreement among the three observers also was present; the interobserver reliability coefficients ranged from 0.81 to 0.92. The measurement of lumbar lordosis is reproducible and reliable if the technique is specified and one accepts 10 degrees as acceptable variation. Factors that affect the reproducibility of measurement include end vertebra selection (especially with transitional segments) and vertebral endplate architecture.