Spinal geometry and paresthesia coverage in spinal cord stimulation.

Objective. To test the following hypotheses, based on computer modeling studies of spinal cord stimulation, by the analysis of data from chronic pain patients: I. the probability-of-paresthesia in a dermatome is highest when the cathode is placed at the corresponding segmental level; II. variation of the rostrocaudal position of the cathode in the lower cervical/high thoracic region results in less variation of the probability-of-paresthesia in a dermatome than stimulation in more caudal regions; III. when stimulating in the midthoracic region, the probability-of-paresthesia in a dermatome is low in comparison to other regions when the cathode is not at the corresponding segmental level. Method. The probability-of-paresthesia in 16 body segments as a function of the rostrocaudal position of the cathode was analyzed from the paresthesia coverage with 3,897 bipolar and unipolar combinations from 106 chronic pain patients. Results. The distributions of the probability-of-paresthesia in the upper and lower limb are in accordance with the hypotheses, but different distributions were found in all trunk areas. Conclusion. The success to be expected from spinal cord stimulation in chronic pain management is inversely related to the thickness of the dorsal cerebrospinal fluid layer at the cathode level. Therefore, preoperative measurement from transverse images can be helpful as a predictor for success. 1998 Blackwell Science, Inc.