The physiologic meaning of the respiratory index in various types of critical illness.

Seven hundred and sixty cardiorespiratory studies, from 151 critically ill or high-risk general surgical patients with extrapulmonary (S) or pulmonary (P) sepsis, cirrhotic liver disease (L), or cardiac failure (C), were analyzed to assess the determinants of a simple, easily obtained measure of respiratory oxygen exchange, the respiratory index (RI). The pattern of cardiorespiratory abnormalities was studied and correlated with the change in RI. The most important relations were with shunt (QS/QT), mixed venous O2 (PvO2), VD, and VE. Higher F1O2 and positive end-expiratory pressure (PEEP) were needed as the RI rose, indicating a greater severity of illness. Regression analysis of all types of critically ill patients and surgical controls showed that QS/QT, PvO2, and F1O2 interacted together to explain most of the variability of the RI. The regressions in each homogeneous patient disease category were all highly significant (p less than .0001) but had somewhat similar coefficients and explained the variability in RI to different degrees. The data suggest that patients with extrapulmonary sepsis or cirrhotic liver disease have an increase in RI (over that in controls) primarily due to a large increase in CI at the high QS/QT caused by the ventilation/perfusion (VA/QT) maldistribution characteristic of these diseases. However, patients with P or C have a disproportionate rise in RI at any given QS/QT compared to that in high-flow states alone, suggesting in P a direct alveolar limitation of oxygen exchange over and above any level of VA/QT mismatching, and suggesting in C a disproportionate decrease in PvO2 that magnifies the QS/QT effect even though VA/QT is more uniform.