330 PHYSIOLOGY OF INDUCED HYPOTHERMIA 



for hypothermia induction. Analysis of the data after completion of the experi- 

 ments revealed that a spontaneous separation into distinct subgroups had occurred, 

 the distinction being evident only at low temperatures and at terminus. The ap- 

 parently related phenomena upon which the separation is based are: (a) the degree 

 of threshold decline with heart temperature, and (b) temperature and mode of 

 death, i.e., ventricular fibrillation or asystole. The upper curve of figure 2 is a plot 

 of the means of the thresholds of three dogs (subgroup A) all of which cooled to 

 16° to 18° before death in asystole. The numbers at each plotted point are standard 

 deviations of the means, which become progressively smaller at the lower tempera- 

 tures. Curve B is a similar plot for the 12 dogs of subgroup B, 11 of which suc- 

 cumbed to ventricular fibrillation between 20° and 26° C. Other interesting points 

 in relation to these data are: (a) the difference between the two subgroups can be 

 demonstrated statistically only after a temperature of 28° is reached, the difference 

 being highly significant at 26° (P<0.01), (b) there is no overlap of the ranges of 

 the two subgroups at 26° or lower, (c ) the range of thresholds of the group B dogs 

 at 24° and lower is 0.27-1.4 ma. with a mean of 1.16, and although such low 

 thresholds may (and did) obtain among both subgroups above 30°, fibrillation be- 

 came associated therewith only after temperature had declined at least to 26°. It 

 would appear from this and many other observations that low thresholds above 28° 

 are not of prognostic importance in acute hypothermia, but become so at lower 

 temperatures. 



Similar measurements were made on a second series of 13 pentobarl)italized dogs 

 in another connection (fig. 3) with practically identical results. And a third series 

 of 15 experiments performed under thiopental anesthesia differed only c^uantita- 

 tively. In the latter series there were five dogs in subgroup A (high preterminal 

 thresholds with asystolic deaths) and 10 in subgroup B. Ventricular fibrillation 

 overtook seven of the latter between 23° and 19°. The other three survived to 15° 

 to 18° without terminal fibrillation. This lower ratio of fil)rillation to asystole under 

 thiopental is in accord with previous observation, but an explanation for the better 

 showing of thiopentalized animals awaits investigation. 



In summary the data from 43 barbitalized dogs, respiring spontaneously to 25° 

 during cooling, indicates that when a high diastolic threshold is maintained dur- 

 ing hypothermia survival is assured to lowest temperatures and death will be 

 asystolic (11 dogs). Failure to maintain or regain a near-normal threshold during 

 hypothermia produces almost certain fibrillation above 19° (27 of 32). This is not 

 to say that the intimate mechanism inducing fibrillation acts necessarily by way of 

 the low diastolic threshold. The low diastolic threshold is jirobably an index of 

 change elsewhere in the cycle of a nature which predisposes to fil)rillati()n. 



Ventricular excitability in relation to hypothermia and pH. The ({uestion 

 may now properly be asked: Are the effects on threshold as observed in hypo- 

 thermia due to temperature per sc, or are they related only indirectly to tem]iera- 

 ture and more directly to other changes induced by low temperature? Since such 

 factors as hypoxia, arterial blood pressure, central venous pressure, autonomic 

 nervous activity, etc., had apparently been ruled out** it became necessary to seek 

 elsewhere for possible primary causes for the increased excital)ility and fibrillation. 

 A report ])y Swan ct al.'* that hv])erventilation during the course of hypotliernna 



