The SYNTAX score revisited: A reassessment of the SYNTAX score reproducibility

OBJECTIVES: To reassess the reproducibility of the SYNTAX score. BACKGROUND: The SYNTAX score appears to have an important role to play in the evaluation of patients with complex coronary artery disease undergoing revascularisation. However, the calculation of the SYNTAX score relies on the subjective assessment of lesions using coronary angiography, and therefore is subject to intra-and inter-observer variability. METHODS: The SYNTAX score was calculated in 100 patients randomly selected from the SYNTAX trial, on two occasions 8 weeks apart, by a team made up of three interventional cardiologists. The weighted kappa values were compared with values obtained 1 year previously, when core lab analysts assessed the intra-observer reproducibility amongst the same patient cohort. RESULTS: The mean +/- standard deviation difference in SYNTAX score was 2.1 +/- 7.6. The respective weighted kappa values for the number of lesions, bifurcation lesions, ostial lesions, and total occlusions were 0.62, 0.36, 0.66, and 0.91 compared with 0.59, 0.41, 0.63, and 0.82 in the previous core lab assessment. The weighted kappa for the intra-observer reproducibility of the SYNTAX score grouped into deciles was 0.54, and according to the terciles < or = 22, >22-< or = 32, >32 was 0.51 both indicating a moderate level of agreement beyond the level of chance. In the previous assessment, the comparative kappa values were 0.45 and 0.53. CONCLUSIONS: The SYNTAX score has moderate intra-observer reproducibility when assessed by a team of three interventional cardiologists, which is consistent with a prior evaluation performed by core lab analysts. The scoring of bifurcation lesions remains the main source of inconsistency.