Right heart strain assessment on CTPA following acute pulmonary embolism: Interobserver variability between expert radiologists and physicians

Background: Accuracy of right heart strain (RHS) measured on computed tomography pulmonary angiogram (CTPA) scans by non-radiologists is unknown. We assessed inter-observer variability of four RHS features and determined the accuracy of measurements by respiratory physicians. Method: 1560 consecutive patients with acute PE were identified, and those who had a CTPA and an echocardiogram within 24-h included. CTPAs were independently scored prospectively by two radiologists, two thoracic physicians and a specialist registrar. Inter-observer variability was assessed, and diagnostic accuracy compared to echocardiography. Results: 182 patients (median age 62.8 years, IQR 49.8-71.5) with acute PE (7.7% high-risk, 40.7% intermediate high-risk, 31.3% intermediate low-risk and 20.3% low-risk) were included. Right ventricle to left ventricle diameter ratio (RV:LV) measurement had low inter-observer variability among the radiologists and non-radiologists with interclass correlation coefficient (ICC) of 0.95 (95%CI 0.92-0.97) and 0.96 (95%CI 0.94-0.97) respectively. RV:LV ratio had high diagnostic accuracy compared to RV dilatation on echocardiography (AUC 0.89, 95%CI 0.84-0.94 for radiologists and AUC 0.84, 95%CI 0.77-0.90 for non-radiologists). Main pulmonary artery to ascending aorta diameter ratio (MPA:Ao) measurement also had excellent agreement amongst the radiologists and non-radiologists (ICC 0.93 (95%CI 0.88-0.96) and 0.92 (95%CI 0.81-0.96) respectively). Significant variability was seen in the assessment of subjective features of RHS (leftward bowing of interventricular septum and contrast reflux into inferior vena cava) amongst the non-radiologists. Conclusion: RV:LV and MPA:Ao diameter ratios on CTPA measured by non-radiologists have low inter-observer variability and good agreement with radiologists, and can be reliably used where an expert report is unavailable.