Electroencephalographic activity and serum and cerebrospinal fluid pentobarbital levels in determining the therapeutic end point during barbiturate coma.

Controversy exists regarding the optimal means for monitoring the patient receiving pentobarbital therapy during medical coma. Serum pentobarbital levels have been used traditionally to gauge cerebral penetration and efficacy of the drug. These peripheral levels have been assumed to reflect pentobarbital concentrations in the cerebrospinal fluid (CSF) and, therefore, the physiological effect on the central nervous system. To determine the relative accuracy of serum versus CSF pentobarbital levels, continuous electroencephalographic (EEG) monitoring in 10 consecutive patients was studied prospectively. Each patient received pentobarbital therapy for cerebral protection in the face of a traumatic injury. Simultaneous serum and CSF pentobarbital levels were obtained 1) before and after the initial barbiturate bolus, 2) every 12 hours during constant infusion therapy, and 3) before and after subsequent boluses necessary because of elevated intracranial pressure (ICP) (ICP greater than 15 mm Hg) or loss of burst suppression by continuous EEG monitoring (defined as greater than five bursts per minute). ICP and relevant clinical events were recorded hourly. Serum and CSF levels ranged from 33 to 74 mg/L and 4 to 54 mg/L, respectively. There was poor correlation between serum and CSF pentobarbital levels at any given time, although patients remained in burst suppression 73% of the time during their therapy. The EEG monitoring not only provided dynamic physiological monitoring, but it also permitted the lowest pentobarbital dose to maintain burst suppression for a specific patient's metabolism, reducing the likelihood of toxicity. In conclusion, CSF pentobarbital levels are of no greater accuracy than serum pentobarbital levels in predicting physiological effect.(ABSTRACT TRUNCATED AT 250 WORDS)