Construction and verification of the failure risk nomogram model of coronary angiography via right radial approach

ObjectiveTo investigate the risk factors of failure of coronary angiography(CAG) via right radial approach, and establish a risk nomogram model.

MethodsNinety-two hospitalized patients with CAG failure via the right radial approach were selected as the failure group using the retrospective case-control study, and 276 hospitalized patients with CAG success via the right radial approach were selected as the control group in a ratio of 1:3.The clinical data of two groups were collected.The possible risk factors of CAG failure via right radial approach were analyzed by univariate and multivariate logistic regression, and the risk factors were incorporated into R software to establish and verify the nomogram prediction model of CAG failure via right radial approach.

ResultsThe results of multivariate logistic regression analysis showed that the female, smoking, radial artery spasm, ratio of radial artery diameter to sheath diameter ≤ 1, operative level < 300 patients per year and LDL-C were the independent risk factors of CAG via right radial approach (P < 0.01).The results of multi-factor analysis were visualized, the risk nomogram model of CAG failure via right radial artery approach was constructed, and the model was validated using RStudio software.The results of ROC curve showed that the area under the curve of model in predicting the risk of CAG failure via right radial approach was 0.816(95%CI: 0.764-0.868), the sensitivity was 0.804, and the specificity was 0.710.The results of calibration curve analysis show that the prediction probability of nomogram prediction model was good agree with the actual occurrence probability, and the mean absolute error of calibration curve was 0.012.Hosmer-Lemeshow test showed that there was no overfitting phenomenon in the model (χ²=10.18, P>0.05).

ConclusionsThe female, smoking, radial artery spasm, radial artery diameter/sheath diameter ≤ 1, operator's operating level < 300 patients per year and LDL-C are independent risk factors of CAG failure via right radial approach.Based on the above risk factors, the risk nomogram prediction model of CAG failure via right radial approach is established, which has good differentiation, calibration and prediction value.