Lisette van Dam

Chapter 10 158 largest transverse diameter. The total thrombus load was assessed by using the CT obstruction index according Qanadli et al. 16 Definitions Acute PE was defined as at least one filling defect in the pulmonary artery tree on CTPA. 17-19 Pleural chest pain was defined as sharp chest pain that worsens during breathing. Non-pleural chest pain was defined as pressure on or squeezing sensation in the chest. PE-related death was defined as objectively confirmed clinically severe PE before death in the absence of an alternative diagnosis. 20 Statistical analysis Baseline characteristics are described as mean with standard deviation (SD) or median with interquartile range (IQR). To evaluate the correlation between PDS to clinical symptoms and adverse outcomes, the difference between the mean PDS with corresponding 95% confidence interval (95%CI) in patients with versus without chest pain, dyspnoea and haemoptysis and adverse short-term outcome was calculated. To evaluate the agreement in PDS scoring between the two reviewers the mean difference between PDS assessed by reviewer 1 and 2 was calculated. The added predictive value of PDS to CTPA assessment for ICU admission, reperfusion therapy and PE-related death was assessed by comparing two prediction models. In the first prediction model CTPA parameters including RV/LV ratio, pulmonary artery trunk diameter and total thrombus load were included. In the second prediction model PDS assessment was added to these CTPA parameters. Multivariable logistic regression analysis and the likelihood- ratio test were performed to assess the predictive value of the two models for ICU admission, reperfusion therapy and PE-related death andwhether PDS assessment significantly improved the predictive value of the model. Additionally, to quantify the performance of the prediction models, we determined the discrimination and calibration. Discrimination refers to the ability to discriminate between those with and those without the outcome and calibration to the agreement between observed outcomes and predictions. Discrimination was expressed with the concordance (c) statistic, by calculating the area under the receiver operating characteristic curve (AUC) with a 95%CI, with discrimination considered perfect if AUC=1, good if AUC > 0.8, moderate if AUC 0.6–0.8, poor if AUC < 0.6, and no better