Relationship between central venous pressure and bioimpedance vector analysis in critically ill patients

To assess the relationship between central venous pressure values and bioelectrical impedance vector analysis (BIVA), which may be used as complementary methods in the bedside monitoring of fluid status.

Cross-sectional evaluation of a consecutive sample.

Intensive care unit of a university hospital.

One hundred and twenty-one consecutive Caucasian, adult patients of either gender, for whom routine central venous pressure measurements were available.

None.

Central venous pressure values and impedance vector components (i.e., resistance and reactance) were determined simultaneously. Total body water predictions were obtained from regression equations according to either conventional bioimpedance analysis or anthropometry (Watson and Hume formulas). Variability of total body water predictions was unacceptable for clinical purposes. Central venous pressure values significantly and inversely correlated with individual impedance vector components (r2 = .28 and r2 = .27 with resistance and reactance, respectively), and with both vector components together (R2 = .31). Patients were classified in three groups according to their central venous pressure value: low (0 to 3 mm Hg); medium (4 to 12 mm Hg); and high (13 to 20 mm Hg). Three BIVA patterns were considered: vectors within the target (reference) 75% tolerance ellipse (normal tissue hydration); long vectors out of the upper pole of the target (dehydration); and short vectors out of the lower pole of the target (fluid overload). The agreement between BIVA and central venous pressure indications was good in the high central venous pressure group (93% short vectors), moderate in the medium central venous pressure group (35% normal vectors), and poor in low central venous pressure group (10% long vectors).

Central venous pressure values correlated with direct impedance measurements more than with total body water predictions. Whereas central venous pressure values >12 mm Hg were associated with shorter impedance vectors in 93% of patients, indicating fluid overload, central venous pressure values <3 mm Hg were associated with long impedance vectors in only 10% of patients, indicating tissue dehydration. The combined evaluation of intensive care unit patients by BIVA and central venous pressure may be useful in therapy planning, particularly in those with low central venous pressure in whom reduced, preserved, or increased tissue fluid content can be detected by BIVA.