In Reply:- Davidson and Ginosar have made some important comments regarding our study, and we appreciate the opportunity to reply. First they point out that temperature is an indirect method to assess the development of a sympathectomy. Therefore they infer that we cannot assume that the presence or absence of temperature changes indicates the development or absence of a sympathectomy. We agree; temperature is an indirect method of assessing the development of a sympathectomy. However, it is an effective and accepted method. In the article by Matsukawa et al., [1]to which Davidson and Ginosar refer, temperature was used to assess whether a sympathectomy occurred. Additionally we used more strict criteria than Matsukawa et al. for determining baseline sympathetic tone. They assumed that a calf-minus-toe index of 0 [degree sign]C indicated the presence of significant baseline sympathetic tone, whereas we required a gradient of 4 [degree sign]C. We maintain that in this group of patients we would have measured some temperature changes had a significant sympathectomy had been induced by intrathecal sufentanil.
Davidson and Ginosar also point out that 6 of 12 subjects in Matsukawa's study did not develop a complete sympathectomy. This is probably not a reflection of methodologic problems with temperature changes failing to demonstrate the presence of a sympathectomy. We doubt that a sympathectomy actually occurred in these subjects and would not have expected temperature changes to occur in all patients. Rather this demonstrates that an epidural block induced with 1% lidocaine to T10 does not consistently block the transmission of all sympathetic stimuli to the lower extremities. Blocks may have been segmental, patchy, or some of the sympathetic control of the lower extremities may derive from above T10.
The above argument is refuted by Davidson and Ginosar, who refer to a paper by Stevens et al., [2]stating that there is no difference between the sympathetic block induced by a spinal and an epidural anesthetic. However, Stevens' paper deals with a different phenomena than the one we are discussing. The authors specifically attempted to induce a complete sympathectomy that would ablate adrenal and cardiovascular responses to a cold pressor stimulus applied in the upper extremity. In contrast Matsukawa et al. [1]and our group were looking at changes in sympathetic tone in the dermatomal levels affected by light, analgesic blocks.
Davidson and Ginosar also point out that core hypothermia in the bupivacaine group was not necessarily caused by redistribution. They claim that it could have been caused by losses to the environment (intravenous fluid, exposure in the operating room, and so on). We believe this is incorrect for several reasons:
1. The core temperature cooled, whereas the periphery warmed up. Had there not been a sympathectomy that resulted in redistribution, then the periphery should also have cooled.
2. The spinal anesthetic was induced after fluid loading. Any changes in core body temperature resulting from the fluid loading had already occurred before induction of spinal anesthesia.
3. The changes in core and peripheral temperatures began immediately after spinal anesthesia induction and were well established before significant environmental losses could have occurred.
Davidson and Ginosar also criticize our choice of a control group (cesarean section patients), claiming they were dissimilar from our study group (laboring women). We appreciated that these groups are different, although we wanted a control group that we were sure would develop a sympathectomy so as to avoid equivocal results. Therefore we chose cesarean section patients as our control group. None of our results can be explained logically by baseline differences between the groups. We maintain that differences between the group did not affect our results or conclusions.
Edward T. Riley, M.D.
Sheila E. Cohen
Department of Anesthesia; Stanford University School of Medicine; Stanford, California
