Added apparatus dead space in the N2-closing volume test: - a refinement or a bias

14 subjects without symptoms of cardiopulmonary disease underwent paired closing volume determinations by the modified single breath nitrogen test. Measurements were performed with an apparatus dead space of approximately 40 ml and approximately 350 ml. A rise in closing volume percent for all subjects was found in measurements with dead space of 350 ml. This Difference was statistically significant at the 0.001 level for the group as a whole. No change in the accuracy of the method, expressed as the standard deviation of each subjects values with and without added dead space cojld be demonstrated. Linear regression analysis of the closing volume percent values measured with added dead space as a function of closing volume percent values measured without added dead space showed the expected highly significant positive correlation. However the regression line was not parallel to line of identity, indicating that the higher the initial value of closing volume percent the higher rise after addition of apparatus dead space. With reference to characterizing a person as normal or abnormal this could be considered a refinement of the method.