Formetric rasterstereography: a new perspective. Reply to Padulo and Ardigó

Abstract

To the Editor, We would like to thank you for the opportunity to comment on the letter to the editor from Drs. Padulo and Ardigo regarding our article entitled "Influence of non-traumatic thoracic and lumbar vertebral fractures (VFs) on sagittal spine alignment assessed by radiation-free spinometry" recently published [1]. In this study, we introduced a radiation-free spinometry as a new method to "raise suspicion of new VFs radiation-free and indicate proper diagnostics, such as radiographs,MRI, or CT." Drs. Padulo and Ardigo question (1) the validity of radiationfree spinometry to assess spine parameters and (2) that spinometry showed better predictive accuracy than historical height loss (HHL) to predict vertebral fractures. Regarding the specific comments, we would like to reply as follows: (1) By questioning the validity of radiation-free spinometry compared to conventional x-rays, Drs. Padulo and Ardigo argue that our data should be viewed with caution. They implicate that radiation-free spinometry is not a sufficiently accurate device to be used in patient care. This is hard to understand, as the primary aim of our study was not to validate radiation-free spinometry but "to investigate the accuracy of spinometry to predict VFs by measuring [thoracic kyphosis] and [lumbar lordosis] based on different fracture locations and fracture grades." No data concerning the correct assessment of thoracic kyphosis or lumbar lordosis by radiation-free spinometry were presented in the study. The letter’s authors support their claim by reporting a low accuracy of the assessment of the Cobb angle and differences of 44 % between radiation-free spinometry and conventional radiographs analyzed by the Raimondi method by interpreting data from another study group [8]. In fact, compared to radiographs, Hackenberg et al. [2, 3] and Drerup and Hierholzer [4] performed validation studies based on a best-fit superimposition and differences between the lateral deviation and rotation curves of radiation-free spinometry and digitized x-rays [5, 6]. This approach takes local measurement shortcomings into account acknowledging that isolated segmental radiographic parameters might be erroneous. Given that the tilt of the single vertebrae cannot exactly be assessed it is widely accepted that the accuracy of the Cobb angle assessed by radiation-free spinometry is limited and not automatically given [2–4]. Although a comparison of radiation-free spinometry with, e.g., radiographs using Bland-Altman plots would be an adequate approach for its validation, we again want to emphasize that this was not the aim of our previous study. From our point of view, the letter’s authors rather miss the opportunity to present data or even Bland-Altman analyses of their own [7, 8]. In fact, due to the previously reported excellent interand intratester reliability of the assessment of thoracic kyphosis (intertester Cronbach’s α=0.979; intratester Cronbach’s α=0.920) and lumbar lordosis (intertester Cronbach’s α=0.992; intratester Cronbach’s α=0.955) with radiation-free spinometry, we feel confident in our results and conclusions [9]. Second, their claim that patients "deserve [...] reduced exposure to ionizing radiation" questioning overall application of radiation-free spinometry in patient care should be viewed with caution. We suggested that the combined assessment of thoracic kyphosis and lumbar lordosis by radiation-free spinometry could help raise suspicion for VFs und The letter referred to here can be found at DOI 10.1007/s00198-0142621-9.