Influence of framerate in the accuracy of detecting oropharyngeal aspiration in paediatric videofluoroscopic swallow study - An observational study

Abstract

Introduction The videofluoroscopic swallow study (VFSS), currently considered the gold standard for assessing aspiration in children, requires radiation. Adhering to the 'as low as reasonably achievable' (ALARA)' principal is crucial in minimising radiation dosage while obtaining accurate diagnostic information, especially in children. Current international adult VFSS guidelines recommend a capture rate of 30 frames per second (fps). Higher capture rates increases radiation yet there is limited evidence on best practice x-ray capture rates in children, particularly on thin fluid consistency - the fastest viscosity with the highest potential for missed aspiration on 15fps. We aimed to determine if image acquisition at 30fps versus 15fps for paediatric VFSS when assessing thin fluid consistency alters the accuracy of detecting aspiration. Material and Methods Seventeen speech language pathologists (SLPs) across Australia and New Zealand with > 2 years paediatric dysphagia experience were recruited. They rated 76 VFSS recordings of infants and children drinking thin fluids for the presence or absence of aspiration of each VFSS recording on 2 separate occasions electronically in a blinded manner. Each VFSS recording was randomly presented at 15 versus 30 fps. An experienced SLP and paediatric radiologist used the penetration-aspiration scale to rate aspiration to form the source of truth for binary ratings (present/absent). Areas under receiver operating characteristic curves (aROCs) were used to compare the accuracy of aspiration ratings at 15 versus 30 fps. The intraclass correlation coefficient was used to examine rater reliability. Results The accuracy for detecting aspiration is similar at 15fps (aROC = 0.97; 95%CI 0.96-0.97) and 30fps (aROC = 0.96; 95%CI 0.96-0.97). Good inter-rater (ICC = 0.82; 95%CI: 0.72-0.89) and intra-rater reliability among the raters (ICC = 0.89; 95%CI 0.82-0.93) was found. Discussion/Conclusion Using 15fps in paediatric VFSS when assessing for thin fluid consistency aspiration provides a similar detection rate to using 30fps. As 15fps would have a lower radiation dose than 30 fps, we recommend using 15fps when undertaking VFSS in children.