Induced hypothermia for infants with hypoxic- ischemic encephalopathy using a servo-controlled fan: an exploratory pilot study.

Abstract

Several trials suggest that hypothermia is beneficial in selected infants with hypoxic-ischemic encephalopathy. However, the cooling methods used required repeated interventions and were either expensive or reported significant temperature variation. The objective of this pilot study was to describe the use, efficacy, and physiologic impact of an inexpensive servo-controlled cooling fan blowing room-temperature air. A servo-controlled fan was manufactured and used to cool 10 infants with hypoxic-ischemic encephalopathy to a rectal temperature of 33 degrees C to 34 degrees C. The infants were sedated with phenobarbital, but clonidine was administered to some infants if shivering or discomfort occurred. A servo-controlled radiant warmer was used simultaneously with the fan to prevent overcooling. The settings used on the fan and radiant warmer differed slightly between some infants as the technique evolved. A rectal temperature of 34 degrees C was achieved in a median time of 58 minutes. Overcooling did not occur, and the mean temperature during cooling was 33.6 degrees C +/- 0.2 degrees C. Inspired oxygen requirements increased in 6 infants, and 5 infants required inotropic support during cooling, but this was progressively reduced after 1 to 2 days. Dehydration did not occur. Five infants shivered when faster fan speeds were used, but 4 of the 5 infants had hypomagnesemia. Shivering was controlled with clonidine in 4 infants, but 1 infant required morphine. Servo-controlled fan cooling with room-temperature air, combined with servo-controlled radiant warming, was an effective, simple, and safe method of inducing and maintaining rectal temperatures of 33 degrees C to 34 degrees C in sedated infants with hypoxic-ischemic encephalopathy. After induction of hypothermia, a low fan speed facilitated accurate temperature control, and warmer-controlled rewarming at 0.2 degrees C increments every 30 minutes resulted in more appropriate rewarming than when 0.5 degrees C increments every hour were used.