Pj. Dillon et al., Comparison of superior vena caval and femoroiliac venous pressure measurements during normal and inverse ratio ventilation, CRIT CARE M, 29(1), 2001, pp. 37-39
Objective: The aim of this study was to estimate the agreement between supe
rior vena caval pressure (SVCP) and femoroiliac venous pressure (FIVP) meas
urements by using short (<20 cm) femoral catheters commonly used in an adul
t intensive care unit. In addition, the effects of two modes of Ventilation
on agreement were assessed.
Design: Measurements of central venous pressure were recorded from both sit
es by using the same pressure transducer connected to the catheters via a t
hree-way stopcock. SVCP and FIVP were recorded at 5-min intervals for 40 mi
ns with the patient in the supine position. Recordings were taken from vent
ilated patients during a randomized crossover sequence of normal and invers
e ratio Ventilation (IRV). Analyses included Pearson's correlation (r), int
raclass correlation (r(l)), Bland-Altman plots, and repeated measures analy
sis of variance with crossover tests for period and period-treatment intera
ctions.
Setting: Adult intensive care unit.
Patients: Adult intensive care patients.
Measurements: Central venous pressure.
Results: Twenty-two patients were enrolled in the study, giving 162 paired
measurements; r was .97 (p < .0001), and r(i) was .96. The bias for SVCP-FI
VP measurements was -0.75 mm Hg (95% confidence interval = -1.31 to -0.18),
with 95% limits of agreement of -3.30 to 1.81 mm Hg. Seventeen patients we
re suitable for randomization to normal ratio ventilation and IRV. IRV sign
ificantly increased SVCP and FIVP (p < .002). Tests for the effect of mode
of ventilation on agreement (p = .36), for period (p = .26), and for period
-treatment interaction (p = .84) were not significant.
Conclusion: The study showed excellent overall agreement with acceptable cl
inical agreement for SVCP and FIVP measurements that was not affected by ch
anging the mode of ventilation. IRV significantly increased central venous
pressure measurements from both catheter sites but had no effect on agreeme
nt.