Oxygen content, blood viscosity, and cerebral blood flow.

329 stated that "The model was soon found to be unreliable, perhaps due to failure to document a lack of brain blood flow during the ischemic period and to nonstandardized postischemic supportive care, aspects of peri-ischemic handling that were not appreciated at this time." In fact, the model itself was never unreliable, as the results in our first report showed.' Indeed, in developing the model, we were able to ascertain very easily if cerebral perfusion occurred during ischemia by the sustained increase in intracranial pressure (ICP) after tourniquet inflation. The ability to produce complete global brain ischemia using this method was repeatedly verified by gamma camera scanning with technetium-99m and also, as mentioned in our initial report, by intra-arterial injection of xenon-133 with external scintillation detection to verify the lack of cerebral perfusion. Therefore, the reliability of the method of producing complete global brain ischemia was never in question. In fact, I would challenge the authors to give us the figures on the number of failed studies due to continued cerebral perfusion despite adequate control of arterial blood pressure and proper placement of the neck tourniquet. The main reasons for failure of complete ischemia in our experience were usually a ruptured tourniquet bladder or failure to adequately control the hyperten-sive response to ischemia. For these reasons, it is usually obvious if cerebral perfusion continues during ischemia, and the monitoring of cerebral perfusion during ischemia really does not make it a "better model." The precision required in the control of postischemic physiological variables, however, is something that we may not have adequately appreciated prior to 1977. Our primary objective in developing this model was to try to control very precisely all postinsult physiological variables to develop a standardized model of complete global brain ischemia in the primate with standardized postinsult care. I believe we succeeded in accomplishing this goal. The degree of precision required in the control of physiological variables, however, we appreciated only later. In evaluating the effects of thiopental in ameliorating ischemic brain damage, we failed to control MAP in the first 5 minutes after ischemia equally between thiopental treated and control monkeys such that MAP at 5 minutes was higher in the thiopental-treated monkeys (about 90 mm Hg) compared with the untreated controls (about 80 mm Hg).3 Mean arterial pressure at 15 minutes after ischemia, however, was similar in both thiopental-treated and untreated control groups. Although it was …