Author + information
RESOLVE score is a novel score system to evaluate the risk of side branch (SB) occlusion when using single stent or provisional 2-stent technique for bifurcation intervention. The RESOLVE score contains 6 independent predictors (2 visual estimation predictors: plaque distribution and main vessel (MV) TIMI flow grade before stenting; 4 QCA predictors: pre-procedural diameter stenosis of bifurcation core, bifurcation angle, diameter ratio between MV/SB and diameter stenosis of SB before MV stenting) of SB occlusion and can help identify patients at risk for SB occlusion during bifurcation intervention with good calibration. However, it is hard to use QCA variables in real-world procedures. In the present study, we aimed to evaluate the capability of risk prediction of side branch occlusion when calculating RESOLVE score completely with visual estimated predictors (V-RESOLVE score).
The present study enrolled all the lesions (1,601 consecutive bifurcation lesions) which have been analyzed in the study for development and validation of RESOLVE score. An independent observer who is blind to previous QCA data performed visual estimation to get V-RESOLVE score. Discrimination and calibration were determined by the C-statistic. Considering the variation of visual estimation, statistical simulation of 30 different observers was performed to further validate the V-RESOLVE score. Simulated observers were set to have an average gap range from -2 to 2 (-2, -1, 0, 1, 2) with the RESOLVE score, and the standard deviation of the V-RESOLVE score was set to range from 0-5 (0, 1, 2, 3, 4, 5). Consistent with RESOLVE score, the high-risk group and non-high-risk group were divided by the lowest score in quartile IV of V-RESOLVE score. The rates of SB occlusion of the two groups were compared.
The consistency between visual estimation and QCA analysis showed fair agreement (Weighted κ range: 0.22-0.44). The C-statistic of V-RESOLVE score was 0.76 (95%CI: 0.71 to 0.80), which is comparable to the C-statistic of RESOLVE score (0.77, 95%CI: 0.71 to 0.80). The net reclassification improvement of the RESOLVE score over the V-RESOLVE score was 15.4% (p=0.11) and the integrated discriminatory index of the RESOLVE score over the V-RESOLVE score was 0.9% (p=0.42). As for the 30 different simulated observers, the C-statistic ranges from 0.65 to 0.77, with all p<0.01. The rate of SB occlusion was significantly different between the high-risk and non-high-risk group decided by V-RESOVLE Score II (Table 1).
|RESOLVE score||V-RESOLVE score||p|
|Non-High Risk Group||Range||0-9||0-11|
|Quartile||I II III||I II III|
|Rate of SB Occlusion||3.35% (39/1163)||4.32% (52/1205)||0.22|
|High Risk Group||Range||10-43||12-43|
|Rate of SB Occlusion||18.04% (79/438)||16.67% (66/396)||0.60|
The V-RESOLVE score can help risk stratification of SB occlusion. There is no significant difference in discrimination and calibration between V-RESOLVE score and RESOLVE score.
Bifurcation, Risk score, Side branch occlusion