Article révisé par les pairs
Résumé : There are presently no means for estimating the risk of seizures carried by a given rate of fluid infusion in the treatment of hypernatremic dehydration. (HD). Fifty seven infants with HD were rehydrated on a 185 ml/kg/24 hr. basis. Among them, 9 (Group I) presented seizures for the first time during treatment; and 16, matched for age, did not convulse (Group II). Subsequent cases were ascribed a 120 ml/kg/24 hr regimen; 8 of these, also matched for age with Group I, constituted Group III. The 3 groups were entirely comparable under many respects, including plasma Na and urea on admission. The fluids were comparable regarding (Na), but their rates of administration were significantly different in the 3 groups: respectively 216, 175 and 123 ml/kg/24 hr for I, II and III. In each group we computed the slopes of the increases in weight over the admission weights as a function of time (α), and the slopes of the decreases in plasma Na with time (β). Despite the differences in the rates of infusion, neither was αI different from αII, nor αII from αIII. By contrast, both βI and βIII differed from βII. Plotting α and β values for each group against infusion rates yielded straight lines with very different slopes, showing that infusion rates affect natremia much more than weight. Thus, it appears that during the treatment of HD, seizures are associated with high infusion rates, which correlate well with rapid falls in natremia; plasma urea on admission is not involved in the seizures; and it may be suggested that the rate of decrease in plasma Na should not exceed 0.5 mEq/l/hr which corresponds to an infusion rate of 150 ml/kg/24 hr, containing 60 mEq/l Na.