Admission pulse pressure and in-hospital mortality in type A acute aortic dissection: result from a Chinese study in stable patients on admission

Figures

• 

  Fig. 1. Flowchart of study patient selection and decision-making. a Patient selection. b Decision-making. PP: pulse pressure; SBP: systolic blood pressure; ECG: electrocardiogram; CPR: cardiopulmonary resuscitation; TBAAD: type B acute aortic dissection; TAAAD: type A acute aortic dissection

• 

  Fig. 2. Relationship between admission pulse pressure and in-hospital all-cause mortality in TAAAD. PP and in-hospital all-cause mortality showed a "J-curve" correlation, with higher rates in the significantly elevated group and reduced group. PP: pulse pressure; TAAAD: type A acute aortic dissection

• 

  Fig. 3. ROC curve of the model for predicting in-hospital all-cause mortality in TAAAD. After plotting the ROC curve, and the AUC of the curve was 0.827 (95% CI 0.774–0.880, P < 0.001). ROC: receiver operator characteristic curve; AUC: area under the curve

• 

  Fig. 4. Nomogram for predicting in-hospital all-cause mortality in TAAAD. Column line plots were drawn to visualize the results of the prediction model. A total of six predictor variables were involved in the composition of the column line plot. Based on the level of influence of each predictor variable in the model, i.e., the magnitude of the regression coefficients, a score was assigned to the level of value of each variable, thus obtaining six individual scores, which were then summed up to obtain the total score. TAAAD: type A acute aortic dissection

• 

  Fig. 5. Calibration curve of the model for predicting in-hospital all-cause mortality in TAAAD. The calibration plot showed that the calibration curve (Apparent line) closely matched the diagonal (Ideal line), indicating good calibration of the predictive model. In addition, internal validation of the predictive model using bootstrap resampling showed that the corrected curve (Bias-corrected line) almost overlapped with the diagonal, suggesting good accuracy of the nomogram model

• 

  Fig. 6. Clinical decision curve of the model for predicting in-hospital all-cause mortality in TAAAD. A DCA curve was drawn based on the constructed CPM. The horizontal axis represented the risk threshold, while the vertical axis showed the net benefit (NB) after accounting for benefits and harms. The DCA results indicated that the risk assessment model provided net benefit for patients within a threshold range of 0.01–0.72