HEMORRHAGIC SHOCK— FRII-DM AN, DAVIDOFF and FINE 371 



5 hours in one and 4', in tlie other, (hn'ini;' whiel: time al)out 40 per cent of the shed 

 hliKid had returned t(i llie aninial>. Tlie rest was then transfused Idic tenijjerature 

 rose alont; with the rise in hl(i<id ])ressure. l)ut the jjressor response was inadecjuale 

 and not sustained IJoth d(i,L;s were dead several hours later. The ])ostniorteni 

 tindinj^s in the two C(»ni])leted experiments were characteristic of irreversil)le 

 hemorrhagic shock. 



The hemodynamic data of hoth experiments are similar to, and are included in, 

 table IV below with the data from experiments with external cooling. 



Because the time recjuired to lower the temperature in the unanesthetized dog in 

 shock was longer than we considered desirable for our i)urposes, because of the 

 unpredictable occiu'rence of ventrictilar fibrillation during .shock as well as before, 

 and because of the undesirable additional load of an extracorporeal circuit, we 

 abandoned the veno-venous cooling method.* 



COOLING BY IMMERSION IN ICE WATER 



This is a much simpler technique than direct cooling of blood and was used in 

 all subsequent experiments. We preferred ether rather than barbiturates as an 

 anesthetic for precooling because, by allowing a short interval for desaturation 

 before shock is induced, virtually all of it is eliminated. Shock dispenses with the 

 need not only for further anesthesia, but also for further cooling because the body 

 temperature continues to fall and remains below 28° C. until a transfusion is given. 



Experimented procedure. Mongrel dogs weighing 15-25 kg. received morphine 

 sulfate (2 mg./kg.J and were lightly anesthetized with ether to prevent shivering. 

 After immersion in ice water for one hour the rectal temperature fell to 28° C. 

 The anesthesia was then stopped, heparin (2 mg./kg. ) was injected intravenously, 

 and the animal was removed from the ice water. Time for desaturation of ether was 

 allowed. The dogs were then bled from a cannulated femoral artery into a bottle 

 elevated so as to maintain the arterial pressure at 30 mm. Hg, as previously 

 described.* The rectal tempei"ature continued to drop during the next half hour, 

 and generally leveled ofif at about 23° C, where it remained for two hours. ( In two 

 of fifteen experiments the rectal temperature dropped to 19° C. ) It then began to 

 rise one-half degree every hour so that at the time of transfusion it ranged between 

 25-28° C. 



The amount of blood entering the elevated reservoir reached a maximum in 

 30 minutes, and remained at that level from one to two hours, after which blood 

 returned to the dog at about 25 ml. per hour, which is very much slower than in 

 uncooled dogs. After 40 per cent of the maximal bleeding volume had spontaneously 

 returned to the animal, or after an arbitrary period of eight hours of hypotension, 

 whichever occurred first, the femoral artery was clamped and the blood remaining 



t Precooling by any method requires anesthesia to prevent shivering. Cooling started during 

 shock does not produce shivering and does not require anesthesia. In the definitive experiments 

 with external cooling described below, we were obliged to use ether for cooling prior to inducing 

 shock. Our experience does not permit a judgment as to the relative merits of the Ross method 

 and external cooling when anesthesia is employed for precooling. But for cooling initiated during 

 shock external coohng proved superior to direct cooling of blood in that ventricular fibrillation, 

 within the temperature limits employed, did not occur and was in general a less hazardous burden 

 on the animal. 



