374 PHYSIOLOGY OF INDUCED HYPOTHERMIA 



lethal factor was involved. The postmortem examinations disclosed none of the 

 normal stigmata of shock, or evidence sufficient to implicate hypothermia as the 

 cause of death, for nothing more than an occasional instance of pulmonary atelec- 

 tasis was found. Since we had already demonstrated that traumatic shock in normo- 

 thermic animals severely impairs the antibacterial defense, and that this impairment 

 persists for at least 48 hours after transfusion in surviving animals, we considered 

 the possibility that infection might account for the late death in cooled animals, in 

 spite of the absence of the customary evidence of infection. In the subsequent series 

 of experiments (Group II, table II) this was investigated by treating the dogs in 

 precisely the same manner as the dogs in Group I except that, as stated, antibiotic 

 was administered at the time of transfusion and twice daily for the three subse- 

 quent days. The result was 100 per cent permanent survival ! 



In uncooled dogs antibiotics given at the time of transfusion proved useless, in 

 contrast to the very consideral^le protection they afforded when given prior to 

 shock. ^ The explanation for the difference in the result due to timing of the anti- 

 biotic is that loss of resistance to Ijacteria develops rapidly during the shock state 

 so that some bacterial toxin is produced in the absence of anti])iotic. At the same 

 time the animal becomes rapidly and increasingly vulnerable to toxin so that if onlv 

 a tiny fraction of the dose that is lethal to the normal animal is produced it may 

 become a fatal dose as the shock continues, since it can meanwhile be neither de- 

 toxified nor excreted." The development of irreversibility to transfusion appears, 

 therefore, to be related to the amount of toxin to which the animal is exposed be- 

 fore an antibiotic to halt further toxin production is administered. Since antibiotic 

 therapy at the time of transfusion is ordinarily useless, but is the sole agent re- 

 sponsible for survival in the precooled animal, the conclusion follows that the pro- 

 tection which precooling confers upon the animal in shock consists in shielding the 

 antil)acterial defense mechanism. That this is true is further demonstrated in ex- 

 periments to be reported elsewhere,* the results of which were briefly as follows : In 

 uncooled dogs transfusion after two hours of shock nearly always results in per- 

 manent recovery. Nevertheless, the antibacterial defense has been damaged because 

 such dogs uniformlv succumb to an intravenous dose of bacteria which the normal 

 dog can readily dispose of.'' This dose of bacteria is fatal whether given during the 

 shock period or anytime within 24 hours after transfusion. ( )n the other hand, if 

 the dogs are precooled, the bacteria are readily eliminated, and the dogs survive. 



In the group of experiments in which cooling was started after shock was in- 

 duced (Group HI), and with antibiotic given as in the Group IT experiments, the 

 protection achieved was far less than by precooling, i.e. a sur\ival rate of 30 per 

 cent in Group III as contrasted to 100 per cent in Group II. From the evidence 

 given this is because during the interval between the induction of shock and the 

 delayed aiiplication of hypothermia souie damage to the antibacterial defense 

 mechanism in c()nse(|uence of deficient ])eri])heral llow occurs and some bacterial 

 toxin is produced. Precooling is sU]u*rior to delated cooling l)ecause it not only 

 minimizes the damage to the defense mechanism during this interval, but ])robably 

 also because it at the same tinu' niinimizes or totally i)revents the ])ro(luction ol 

 bacterial toxins. In this latter respect ])ropliylactic hypothermia achieves the same 



