Observational study in healthy volunteers to define interobserver reliability of ultrasound haemodynamic monitoring techniques performed by trainee doctors

Objectives Bedside ultrasound is increasingly being used to guide fluid management in shocked patients. Little data exist on the inter-rater reliability of techniques used, especially when performed by nonexpert trainee doctors. The primary aim of this study is to measure the inter-rater reliability of five ultrasound techniques commonly used to guide fluid management: inferior vena cava collapsibility index (IVCCI), transthoracic echocardiography (TTE)-derived stroke volumes, ultrasound cardiac output monitor (USCOM) derived stroke volume and carotid artery blood flow and corrected flow time measurements. Methods Two Royal College of Emergency Medicine level one ultrasound-certified emergency medicine trainees performed paired ultrasound measurements on 31 healthy nonpatient volunteers. Inter-rater reliability was assessed through three indices: interclass correlation coefficient (ICC), limits of agreements (LOAs) derived from Band-Altman plots and the proportion of paired scans with absolute differences of less that 15% (defined as agreement). Results TTE-derived measurements performed the best overall, with an LOA of 22%, an ICC of 0.55 and an agreement of 80%. USCOM also performed well, with an LOA of 33%, an ICC of 0.68 and an agreement of 58%. IVCCI and carotid artery-derived measurements performed poorly across all indices. Conclusion TTE-derived measurements showed the highest level of inter-rater reliability and can thus be expected to provide reliable measures over time with different sonographer clinicians. USCOM interobserver reliability was also adequate for clinical use. However, on the basis of inter-reliability measures, IVCCI and carotid artery measurements were found to be inadequate for clinical use.