Influence of diffusion on estimations of protein reflection coefficient by double-indicator method

In isolated perfused organs, vascular protein reflection coefficients (σ) can be calculated from the changes in hematocrit and perfusate protein concentration (C(P)) that occur during edema formation. This technique requires the assumption that transvascular protein flux by diffusion is negligible. To assess diffusion-induced errors in calculations of σ, we derived an expression for C(P) that includes determinants of diffusive protein flux: protein permeability-surface area product (PS), transvascular fluid flux (J), true σ, and transvascular protein concentration. We used this expression to obtain values of C(P) under various experimental conditions and then calculated values of σ (measured σ) for those conditions. Diffusion causes measured σ to be lower than true σ. The diffusion-induced error is larger and potentially substantial when J/PS is low and when true σ is high. Diffusion-induced error is also larger when the amount of edema formation is greater. In recent isolated canine lung experiments where J/PS was ~2.7, diffusion-induced errors in measured σ for albumin would have been ~0.06 (at true σ = 0.5) and ~0.18 (at true σ = 0.9). When J/PS was higher, the potential for diffusion-induced errors was much smaller. We conclude that diffusion causes underestimation of true σ and that the error in measured σ may be substantial when J/PS is <5 and when true σ is >0.5.