Background: Poorly performing diagnostic tests can impact patient safety. Clinical investigations must have good precision and diagnostic accuracy before widespread use in clinical practice. Transient elastography (TE) measures liver stiffness, a surrogate marker of liver fibrosis in adults and children. Studies to evaluate its repeatability and reproducibility (precision) in children are limited. Our aim was to determine (i) the normal range of TE measurements and (ii) the repeatability and reproducibility of TE in healthy children.
Methods: TE was performed in 257 healthy children, of whom 235 (91%, mean age 11.7 years, standard deviation (SD) 2.51, 107 were males (45.5%)) had two valid TE measurements performed, at least 24 h apart, by two operators under similar circumstances. High-quality TE images were obtained for each examination.
Results: The normal range of TE was 2.88-6.52 kPa. The mean difference between paired measurements was 0.044 (SD 0.4). The 95% limits of agreement ranged from -0.8 to +0.76 kPa for repeat measurements. There was a difference of >1 kPa between measurements in 61/235 (25.9%) children. The lack of precision was similar across all age groups.
Conclusions: This study demonstrates that TE does not have acceptable precision in healthy children, because random measurement variation results in the lack of agreement between paired measurements.
Impact: The precision and diagnostic accuracy of a new technology must be determined before it is deployed in children in order to ensure that appropriate clinical decisions are made, and healthcare resources are not wasted. TE is widely used to diagnose liver disease in children without adequate evaluation of the precision (repeatability) of TE either in healthy children or children with liver disease. This study demonstrates that TE does not have adequate precision in children. This study was performed in accordance with methods previously published for children. Refinements to the test protocol, such as duration of fasting or probe size, will have to be evaluated for their impact on precision and accuracy before the test is deployed in research studies or clinical practice.