Methods
The main modification of the setup in the present study compared with that previously reported is the design of the additional resistor. The new resistor (Fig 1) was made of three layers of stainless steel screen, two layers of No. 400 mesh and one layer of No. 300 mesh. The screens were sandwiched between two thin pieces of plastic, which were then glued together (total thickness, 0.36 mm). A circular area of mesh, with a diameter of 1 inch and an area of5.067 sq cm, was exposed to the subjects air flow. The rest of the apparatus consisted of a rotary solenoid valve in which the resistor was mounted, a heated Fleisch-type pneumotachograph (HP 21071), a differential pressure transducer (Validyne MP45-14-871 ±2 cm HaO), a stimulator (Grass SD9) to standardize the duration of the addition of Rk (0.1 s), and a polygraph to record flow (Grass 79D). Though a single channel pen recorder may be all that is required to record the necessary data, for this study we digitized the flow signal and used a microcomputer to calculate R„.

The pressure-flow characteristics of the apparatus were measured by simultaneously measuring the flow through the pneumotachograph and the pressure difference from one end of the tubing to the other. Flow was varied continuously from 0 to 1 Us. Five resistors with different densities of wire mesh were studied, along with the multiperforated steel plate used previously. Substituting the wire mesh for the multiperforated plate resulted in increased linearity of pressure-flow relationships (Fig 2). To determine which wire mesh resistor to use in the present study, we measured the R„ of five subjects using each density of mesh. There was no systematic variation of mean resistance with We therefore decided to use the resistor previously mentioned after considering two conflicting factors. These were (1) the feet that the coefficient of variation of the measured resistances decreased with increasing Rk (Fig 3) and (2) the need to use the lowest possible R* to minimize intrathoracic gas compression and any resulting change in driving pressure. The resistor with Rk=2.17 seemed to offer both a relatively low Rk and an acceptable amount of variation.

Figure 1. New resistor used in the present study for determining total respiratory resistance by partial airway occlusion.

Figure 1. New resistor used in the present study for determining total respiratory resistance by partial airway occlusion.

Figure 2. Pressure-flow relationship of the whole system in the present study. “Open” means shutter open and all other curves recorded with shutter closed. Numbers at end of the curves represent the layers of the stainless steel screen: 4,4>4>4 connotes screen with four layers of No. 400 mesh; 4,4,3 means two layers of No. 400 mesh and one layer of No. 300 mesh, etc; mp refers to multiperforated steel plate.

Figure 2. Pressure-flow relationship of the whole system in the present study. “Open” means shutter open and all other curves recorded with shutter closed. Numbers at end of the curves represent the layers of the stainless steel screen: 4,4>4>4 connotes screen with four layers of No. 400 mesh; 4,4,3 means two layers of No. 400 mesh and one layer of No. 300 mesh, etc; mp refers to multiperforated steel plate.

Figure 3. Coefficient of variation of resistance for five subjects using five different densities of mesh. Numbers along curves refer to layers of mesh, as in Figure 2.

Figure 3. Coefficient of variation of resistance for five subjects using five different densities of mesh. Numbers along curves refer to layers of mesh, as in Figure 2.