To the Editor: Schramm et al. (2) present a novel and apparently effective approach to hemisphere deafferentation in patients requiring the procedure for intractable seizures related to unilateral porencephaly or unilateral cortical atrophy. As the authors state, however, the technique is less useful for patients with increased volume of the affected hemisphere, i.e., unilateral megalencephaly, and is probably less useful in most children with Rasmussen's syndrome who have little or no cortical atrophy early in their disease. The technique presented by Schramm et al. was developed in response to the high morbidity and mortality rates reported by Brian et al.(1) (J. Schramm, personal communication) in the early Johns Hopkins hemispherectomy series. Brian et al. were not members of the Johns Hopkins hemispherectomy team, and no one on our team had either reviewed or been aware of their article before its publication. We regret the misconceptions engendered by that article. The purpose of this letter is twofold. First, we want to counteract the erroneous impression that “the technique described by Brian has been abandoned.” It has not. We have just completed our 57th hemidecorticectomy. Second, we want to correct the incorrect impressions given by the article by Brian et al. regarding the intraoperative complications and mortality of the procedure. The complications had been independently identified and corrected before the article by Brian et al. by changing our anesthesia procedures and changing our anesthesia team to one run by a single pediatric anesthesiologist (AZ). A full report on our current hemispherectomy anesthesia procedures is in preparation, as is a report on the outcomes of series recently presented at the Bethel-Cleveland Clinic Epilepsy Symposium(Bielefeld, Germany, March 23-26, 1995). Since these anesthesia changes have been implemented, we have had no further perioperative mortality, and the sole important morbidity has been a coagulopathy in some small infants. There has been no mortality in our total series and little long-term morbidity. In summary, our hemidecorticectomy technique is alive and well, with no evidence of hemosiderosis and no evidence that hydrocephalus is any more frequent than at institutions using functional or deafferentation techniques. The incidence of seizures after hemidecortication may be less than that after functional hemispherectomy. Although the “best” technique for hemispherectomy remains to be determined, it may partially be determined by the reason for the procedure. The population in the series presented by Schramm et al. is very different from that in our series. At present, the best technique seems to be the one with which the specific epilepsy center has most experience. Whichever technique is used, such operations are big and difficult and should be performed only as part of a major epilepsy surgery program. John M. Freeman Eileen P.G. Vining Benjamin Carson Aaron Zuckerberg Baltimore, Maryland