Most cases of large degrees of subsidence are just error of judgment when examining the x-ray, and in that first study TRS posted it stated that the conclusion was "Our results indicate that no subsidence of clinical interest occurs after external frame removal."
Quoted from the first study:
"There have been very few reports dealing with possible changes in length of the callotasis zone after removal of an external fixator in DO. Fracture or collapse of the regenerated bone rarely occurs, but a small compression of the newly formed bone that cannot be seen on plain radiographs may be more frequent. There have been anecdotal reports of this delayed complication phenomenon (Aldegheri et al. 1989, Paley 1990). To our knowledge Shyam et al. (2009) are the only authors to have published measured values of delayed loss of length or callus subsidence in DO. They found a tibial subsidence (compression) of 4–32 mm, with about half of the 81 lengthened segments in 48 patients showing a subsidence of 1 cm or more. However, we have argued that the authors did not consider the effect of the divergence of the X-ray beam (Gunderson et al. 2010). They made direct measurements on the electronic images without calibration for the change in distance between the tibia and the film before and after removal of the ring fixator. This error has been acknowledged by the authors (Song and Shyam 2010).
The RSA method itself has a high accuracy, and changes in order of 0.01 mm can be detected (Kärrholm et al. 1997). In most reports dealing with RSA, the precision in clinical settings has been found to vary between 0.15 and 0.60 mm (Steen et al. 2001, Kärrholm et al. 1997). These studies have been performed with old calibration cages and manual measurements made on scanned radiographic films. In the present work, the accuracy and precision of the measurements were estimated to be 0.30 mm and 0.43 mm, respectively. By the use of optimized conditions with the newer model of calibration cage (# 43), 2 ceiling-mounted X-ray tubes fired by 1 single button, no scanning of films, and the latest version of software, a better precision would be expected (Börlin et al. 2002, Mäkinen et al. 2004).
The range of subsidence from –0.37 to 0.27 mm between the 2 post-removal examinations can be regarded as a measure of error or precision of our study. These errors are most likely due to practical and functional conditions: e.g. the geometry of the location of the spheres in the bone, the degree to which the spheres are hidden by the metal, and finally, the positioning of the patient at the moment of examination. However, all our measured values were less than the size of the error or less than 1 mm, which indicates that the subsidence of callotases after removal of the external fixator is negligible and of no clinical interest."