The reliability of respiratory inductive plethysmography (RIP) for non-invasive measurement of tidal volume (VT) in VLBW infants has not been reported. We studied 7 intubated preterm infants (BW 834±92g GA 24 to 27 weeks) on the first week of life. Measurements were done on spontaneous breaths while the infants were on low IMV settings. RIP was automatically calibrated by the qualitative diagnostic calibration (QDC) method in supine position while resting quietly. QDC assigns gain factors to the inductance signal obtained from the rib cage and abdomen, the sum (SUM) of which should correlate with VT. Simulataneous flow recording with a pneumotachograph(PNT) was used to compare the VT obtained by the two methods. RIP and flow derived VT(VTRIP and VTPNT) were recorded for a period of 3 hours while the infants were kept in the supine position. A conversion factor was derived from the slope of a linear regression done after QDC with 20 artifact free end inspiratory values of SUM and VTPNT. This factor was subsequently applied to 3 different sets of 20 breaths obtained at 30, 60 and 120 minutes after QDC. On each set, calibration was validated by obtaining the correlation coefficient (r) and the slope between VTPNT and VTRIP. Calibration was considered acceptable if r was >0.85 and the slope was within 0.85 and 1.15(i.e. within 15% of the line of identity). Furthermore, we obtained the number of breaths in which VTRIP was ± 15% of VTPNT for each set of 20 breaths. Table

Table 1 No caption available.
Full size table

We conclude that the reliability of RIP to measure VT in infants<1000g is low and does not depend on the time elapsed after the QDC calibration. This may be explained by the variable contribution of the chest wall to the SUM depending on the degree of chest wall distortion.