Lung water changes after thermal injury. The effects of crystalloid resuscitation and sepsis.

Respiratory failure after thermal injury is common, but the etiologic roles of high volume crystalloid resuscitation, hypoproteinemia, inhalation injury, or sepsis have not been specifically defined in human studies. We used the thermal-green dye double indicator dilution measurement of extravascular lung water (EVLW) to follow daily lung water changes in seven severly burned adult patients, resuscitated with only crystalloid solutions. An average weight gain of 21.3 kg, a 30% increase (p < 0.001), was present two to three days after admission. Admission EVLW for all patients was 7.9 +/- 1.2 ml/kg, (means +/- SD), and EVLW at the time of maximal weight gain was 5.9 +/- 1.4 ml/kg, a 25% decrease (p < 0.05). Admission pulmonary artery wedge pressure (PAWP) was 8 +/- 3 mmHG, which was not significantly different from PAWP of 13 +/- 4 mmHg at the time of maximal weight gain. In the three patients who died of sepsis, their terminal weight averaged 17.8 kg (27%) above their admitting weight (p < 0.01) and EVLW was 26.4 +/- 4.4 ml/kg, a 200% increase (p < 0.02) from admission. Their terminal PAWP averaged 22 +/- 2 mmHg, a 170% increase (p < 0.005). None of these patients had an increase in EVLW until clinical signs of sepsis occurred and the rise in EVLW preceded the rise in PAWP. Calculated mean plasma colloid osmotic pressure (PCOP) on admission was 20.7 +/- 4.9 mmHg; at the time of maximal weight gain, it was 8.6 +/- 1.7 mmHg (p < 0.001). The PCOP-PAWP gradient fell to -4 +/- 4 mmHg (p < 0.001) at the time of maximal weight gain and remained less than +4 mmHg throughout the study period in all patients. We conclude that massive crystalloid resuscitation while maintaining PAWP below 15 mmHg does not cause an increase in EVLW during the first four days after thermal injury. EVLW actually decreases slightly in all patients despite marked weight gain, hypoproteinemia and a negative PCOP-PAWP gradient. EVLW does not correlate with the PCOP-PAWP gradient in either septic or nonseptic periods. Three patients had severe inhalational injury and normal EVLW for the first four postburn days. It therefore appears that significant interstitial edema does not result from inhalational injury. There is also no evidence that thermal injury causes an early increase in pulmonary capillary permeability. The occurrence of sepsis, however, results in rapid accumulation of lung water, without any change in hydrostatic or osmotic forces. This study supports the primary role of sepsis in altering pulmonary capillary permeability with resulting pulmonary edema.