The N2 closing volume test in population studies: sources of variation and reproducibility.

Sources of variation in the nitrogen closing volume test and derived measurements were examined in the results of 13 subjects, each of whom performed 3 sequential trials on 2 occasions 30 to 60 minutes apart, on 2 separate days 1 week apart (156 trials in all). Results were examined to evaluate the relative sensitivity of the various measurements for differences between sujbects and, by implication, their potential value in population studies. Using the ratio of signal (between-subject variance) to noise (within-subject variance) as the criterion, the sensitivities of the ratio of closing volume to vital capacity (CV/VC,%) and the ratio of closing capacity to total lung capacity (CC/TLC,%) were comparable; contrary to expectation, sensitivity decreased rather than increased with side-by-side compared to independent tracing analysis. Comparison of various test schedules showed the greatest sensitivity when the mean of 3 measurements was used, with a single measurement of CV/VC,% being only one-third as sensitive, and a single measurement of CC/TLC,% being one-half as sensitive as the mean. In addition, the validity of the measurements of total lung capacity from nitrogen dilution in a single oxygen breath was confirmed by the demonstration of good agreement with helium dilution values.