Respiratory dysfunction following cardiopulmonary bypass: verification of a non-invasive technique to measure shunt fraction.

Respiratory dysfunction is a well recognized complication of cardiopulmonary bypass. The size of the pulmonary shunt fraction is the best indicator of respiratory dysfunction but its measurement conventionally requires use of a pulmonary artery catheter to measure mixed venous oxygen content. We compared pulmonary shunt fraction, based on a non-invasive technique using a previously described mathematical model, with shunt fraction measured invasively using a pulmonary artery catheter in 22 patients undergoing elective coronary artery surgery. The mean shunt fraction measured by the invasive technique was 19.6 +/- 2.0 (18.8-20.4)% of cardiac output at 24 h (+/- 1 SD and 90% confidence intervals) and 20.9 +/- 2.9 (19.8-22.0)% of cardiac output at 44 h post-surgery. There was good agreement between the two methods of measurement. The mean difference was 0.21 percentage points with 95% confidence interval -0.01 to 0.43. The limits of agreement (-1.17 to 1.59) are small enough to be confident that the non-invasive method can be used to give the same result as that obtained using a pulmonary artery catheter. The values for shunt fractions obtained by the non-invasive technique were 19.7 +/- 2.3 (18.8-20.6)% of cardiac output at 24 h and 20.7 +/- 2.5 (19.7-21.6)% of cardiac output at 44 h. The non-invasive measurement of the shunt fraction provided us with a simple, practical method for following a further ten patients over an extended period of time where prolonged catheterization is impractical.