Effect of positive-end expiratory pressure on accuracy of thermal-dye measurements of lung water

The thermal-dye indicator dilution technique of measuring extravascular lung water (EVLW- TD) correlates well with gravimetric measurement of lung water (EVLW- GR) when no positive-end expiratory pressure (PEEP) is used in both normal and edematous lungs. PEEP is often used to raise arterial oxygen tension in acute respiratory failure. This study was designed to answer the question: Does PEEP have an effect on the accuracy of the EVLW measurement by thermal dye? Sixteen mongrel dogs were anesthetized and intubated. Arterial and PA catheters were placed. They were divided into three groups and ventilated at PEEP levels of 5, 10, or 15 cm H 2O. Fluids were given to minimize decrease in cardiac output with institution of PEEP. They were maintained for 5 hr with measurement of vital signs made hourly and measurement of blood gases and EVLW-TD made at baseline, 1, 3, and 5 hr. After final measurements, gravimetric determination of EVLW was done. Correlation between EVLW TD and EVLW GR remained good when low levels of PEEP were used. At 5 cm H 2O PEEP, EVLW TD was 7.5 ± 0.9 and EVLW GR was 5.4 ± .3. At 10 cm H 2O PEEP, they were 10.0 ± 0.9 and 6.5 ± 0.3. The correlations were 0.87 and 0.97, respectively. However, at 15 cm H 2O PEEP, EVLW TD was 11.4 ± 1.3 and EVLW GR was 7.5 ± 0.6, with a correlation of only 0.59. The correlation between the two techniques seems to break down with higher levels of PEEP in dogs with normal lungs.