Selection of optimal descriptive variables for each system required consideration not only of construct validity, but also of content validity, i.e. the extent to which a candidate variable encompassed the entire spectrum of abnormality for the organ system of interest. Thus, in addition to ascertaining what variables have been employed by others in the past, a set of criteria were developed that characterized the 'ideal' descriptor of organ dysfunction ( Marshall 1.995), and candidate variables were evaluated against these.
Organ dysfunction or failure can be described from one of two perspectives: the physiological perturbation present in a given organ system, or the intervention required to support failing organ function. Only variables that reflected physiological abnormalities were included, because the objective of the score was to describe a biological, rather than a therapeutic, state and because therapeutic decisions vary significantly between centers, thus introducing significant potential for bias.
It was possible to identify acceptable variables for six of the seven organ systems cited in published reviews of multiple organ failure. However, no satisfactory measure of physiological dysfunction could be found for the gastrointestinal system; therefore it was eliminated from further consideration. None of the published descriptors for the cardiovascular system performed well when evaluated against the criteria describing an ideal measure of organ dysfunction. Therefore the pressure-adjusted heart rate (PAR), a physiological measure that corrects for the effects of therapy, was developed:
where CVP is the central venous pressure and MAP is the mean arterial pressure. Increasing degrees of intrinsic cardiovascular dysfunction result in increasing values of the PAR. If the central venous pressure is low (reflecting hypotension resulting from an inadequate circulating volume rather than intrinsic cardiovascular derangement), the value will be similarly low; thus the PAR, like the ratio of PO2 to FiO2, partially corrects for the effects of therapeutic intervention.
Was this article helpful?