Predicting increased fluid requirements during the resuscitation of thermally injured patients

We determined whether factors present soon after burn predict which patients will receive more than 4 mL/kg/% burn during the first 24 hours, and whether total fluid intake during the first 24 hours (VOL) contributes to in-hospital mortality (MORT).

We reviewed the records of patients admitted during 1987-97. The modified Brooke resuscitation formula was used. One hundred four patients met inclusion criteria: total body surface area burned (TBSA) > or = 20%; admission directly from the field; weight > 30 kg; no electric injury, mechanical trauma, or blood transfusions; and survival > or = 24 hours postburn. Eighty-nine records were complete.

Mean TBSA was 43%, mean full-thickness burn size was 21%, mean age was 41 years, mean VOL was 4.9 mL/kg/% burn, and mean lactated Ringer's volume was 4.4 mL/kg/% burn; 53% had inhalation injury. MORT was 25.8%. Mean urine output was 0.77 mL/kg/h. By linear regression, VOL was associated with weight (negatively) and full-thickness burn size (r2 = 0.151). By logistic regression, receipt of over 4 mL/kg/% burn was predicted at admission by weight (negatively) and TBSA; by 24 hours postburn, mechanical ventilation replaced TBSA. With respect to MORT, logistic regression of admission factors yielded a model incorporating TBSA and an age function; by 24 hours postburn, the worst base deficit was added.

Burn size and weight (negatively) were associated with greater VOL. However, a close linear relationship between burn size and VOL was not observed. Mechanical ventilation supplanted TBSA by 24 hours as a predictor of high VOL. Worst base deficit, TBSA, and an age function, but not VOL, were predictors of MORT.