Twenty patients with
delirium tremens (grade 3) and a less severe clinical state (grade 2) were investigated thoroughly from the time of admission until recovery from the acute state. A lumbar puncture was performed in the majority of the patients immediately after admission and then repeated after recovery from the acute state. The cerebrospinal fluid was found to be both macroscopically and microscopically normal, as was the spinal fluid pressure. The
clinical course was without complication, none of the patients were severely dehydrated. All the patients were treated with
barbital, a long acting
barbiturate. The duration of the acute state and the total amount of
drug necessary in the treatment were equal in the two groups of severity. However, patients with proper
delirium tremens needed significantly fewer
barbital doses during the first hours
after treatment was initiated than did patients with a less severe clinical state. The opposite was seen about 12 hours later. These findings are discussed in relation to the high
blood alcohol concentration seen at the time of admission in the majority of the patients with proper
delirium tremens, but not in patients with grade 2. It is concluded that
barbital exerts its effect due to cross-dependence properties with alcohol. The majority of the patients had moderately elevated blood pressure, pulse rate and rectal temperature at the time of admission; these variables were to a great extent normalized within 48 hours after admission. No differences in those physical signs were seen between patients with fully developed
delirium tremens and patients with less severe clinical states. The patients' condition during the acute state was followed by means of a
delirium tremens rating scale. Physical symptoms were similar in various degrees of severity of the clinical condition. 18-24 hours after admission the differences in mental symptoms between patients with grade 3 and patients with grade 2 had disappeared, 48 hours after admission the patients' condition was to a large extent normalized. Methodological problems in using a rating scale in conditions as
delirium tremens are discussed. The results are discussed in relation to aetiology and pathogenesis of
delirium tremens. It is concluded that it may be that a qualitative, and not only a quantitative, difference exists between a severe withdrawal reaction and fully developed
delirium tremens, and a hypothesis about a "point of no return" is suggested.