Reviewer’s comment


understanding spinal biomechanics, for implant development and for the development of spine models. In order to meet these requirements, the anatomy of the human lumbar spine has been the subject of many publications [1, 3, 4, 7, 9, 10]. However, the quality of these publications varies widely. For a high quality anatomical study, an exact measurement method is indispensable. Spinal dimensions have been measured in vitro using fresh cadaveric specimens [3, 4]. Skeletons have also been measured [1, 7]. These skeletons were dried and their dimensions may therefore no longer be identical with those before death. Anatomical dimensions have also been measured in vivo from conventional X-ray films [10]. However, such X-rays project three-dimensional spine onto a two-dimensional film. Due to this projection, a magnification factor arises which is difficult to determine exactly. Besides X-ray films, CT scans have often been evaluated [9, 10] They enable additional measurements to be made in the transverse planes. One of the disadvantages of this method is an overestimation of the cortical thickness caused by the limited resolution of clinical CT systems in combination with a large slice thickness [5]. This explains the large difference between the cortical thickness of 2.6–2.9 mm measured in these investigations and the real values of around 0.29mm [6] and 0.35 mm [8] reported in the literature. The picture archiving and communication system (PACS) has been established in radiology since the early 1990s. Its accuracy has been investigated with sonography, and no statistically significant difference in the error rate for film versus monitor display format could be found [2]. However, its accuracy for determination of dimensions from CT scans has not yet been determined. The quality of an anatomical paper depends not only on an exact measurement method, but also on a large, representative sample. This paper presents geometrical data obtained from a total of 126 patients – more than have been examined in other studies. However, the study was performed with patients suffering from low back pain. Even though patients with visible pathological findings were excluded from the evaluation, particularly the disc height of this patient collective may not be representative of the healthy population. For measurement of the endplate circumference and area, only ten patients, with a distinct disc height, were selected. This is a rather small and specific population. These kinds of papers provide a valuable database for obtaining answers to specific questions arising from practice, such as: “How long should a screw be for anterior cervical plating?” or “Where should a pedicle screw best be placed into the sacrum?”. In summary, this publication presents a wide, but not new, anatomical database. A review of the pertinent literature presenting all known measured dimensions with all applied methods would have been more informative than a new study measuring dimensions that have already been published. Nevertheless, this study may become helpful for clinicians and scientists if interpreted carefully in consideration of the restricted accuracy of measurement and the restrictedly representative sample.

DOI: 10.1007/s005860000151

Cite this paper

@article{Wilke2000ReviewersC, title={Reviewer’s comment}, author={Hans-Joachim Wilke}, journal={European Spine Journal}, year={2000}, volume={9}, pages={249-249} }