Measurement of lumbar lordosis: Inter-rater reliability, minimum detectable change and longitudinal variation

Study Design: Repeated measures design to examine reliability and longitudinal variation of lumbar lordosis measurement. Objectives: To determine the interrater reliability, minimum detectable change (MDC) and longitudinal variation of the Cobb method for measuring lumbar lordosis using standardized rules. Summary of Background Data: The reliability of the 4-line Cobb method for measuring lumbar lordosis was not examined when standardized rules were instituted for drawing the lines. Methods: A random sample of participants was selected from the Pittsburgh clinic of the multicenter Study of Osteoporotic Fractures for radiographic measurement of lumbar lordosis reliability (n = 48) and stability (n = 109). A standardized version of the 4-line Cobb method was used for all measurements of lordosis. The Intraclass Correlation Coefficient (ICC) was used to calculate interrater reliability for lordosis and to measure the stability of this measure over an approximate 2-year-time period. The standard error of measurement and MDC were calculated for lordosis measurement based on the ICC value. Results: The interrater reliability coefficient for lumbar lordosis was in the excellent range (ICC = 0.98; 95% CI: 0.95, 0.99). The MDC based on measurements between raters was 3.90 degrees. The ICC value for the stability, or reliability from time 1 to time 2, of lordosis measurement over time was 0.81 (95% Cl: 0.74, 0.87). Conclusion: This study demonstrates that the 4-line Cobb method can be a highly reliable and precise method for measuring lumbar lordosis if standardized procedures are used. The Cobb method has an MDC that is appropriate for clinical use. Also, there is minimal longitudinal variation in lordosis measurements over a 2-year period.