When the model does not fit the data, the model is wrong.

In Reply:—We appreciate the comments of Drs. Shulman and Robelen, and of Dr. Kitagawa, in emphasizing the potential for neurotoxicity of undiluted 10% procaine in spinal anesthesia, consistent with our case report. We also appreciate the correspondents including references to the work of Watkins and colleagues from early publications which have not been included in modern medical indices, as they deserve wider attention than they have received. One of us had earlier pointed out the relevance of those publications to the question of procaine spinal neurotoxicity, and so did not comment further on them in our case report except to refer to the earlier comment. We note that these references, as well as many other early publications reporting neurotoxicity after procaine spinal anesthesia, were included in the review of Schildt, which we also referenced in the case report, and which has also been undeservedly neglected. Both letters pose the question of why 10% procaine was chosen as the spinal anesthetic in our case report. We cannot answer this because we did not provide the spinal anesthetic, but were consulted after the unfortunate development of the patient’s cauda equina syndrome. Drs. Shulman and Robelen correctly note two anesthesia textbooks which do recommend that procaine not be injected at concentrations higher than 5%. At 10 and 21 yr past their publication dates, it is a matter of semantics whether they would be considered recent texts. We respectfully disagree with the statement of Drs. Shulman and Robelen that “there is no evidence at present that it [5% spinal procaine] is neurotoxic.” The animal study of MacDonald and Watkins cited by Dr. Kitagawa (his reference 3) and Drs. Shulman and Robelen (their reference 7) found that 10% spinal procaine was neurotoxic in 33% of 33 animals tested, but also found that 5% procaine was neurotoxic in 10% of 20 animals tested. This is a higher incidence of spinal neurotoxicity than we are aware of in any animal study testing roughly equipotent single injections of other local anesthetics, although given the limitations in the older report, there is simply not enough data at the present time to compare precisely the potential for neurotoxicity of procaine and lidocaine. Our point is that just because procaine is less likely to cause the syndrome of transient neurologic symptoms, it ought not to be considered an innocuous, risk-free alternative to lidocaine.