Background: Although hypotension commonly accompanies intermittent hemodialysis for acute renal failure (ARF) in the intensive care unit (ICU), little is known about how it may be prevented. Online relative blood volume (RBV) monitoring has been used to reduce hypotension in chronic hemodialysis, but is of unproven benefit in ARF. This study tested the hypothesis that hypotension is predictable using online RBV monitoring in patients dialyzed for ARF in an ICU setting.
Methods: The primary definition of hypotension was mean arterial pressure (MAP) <70 mm Hg; systolic blood pressure (SBP) <100 mm Hg was used as a secondary outcome. Fifty-seven treatments were prospectively studied in twenty consecutive adult patients treated with intermittent hemodialysis. RBV was continuously measured using the Hemoscan device (Gambro).
Results: Hypotension occurred in 30% of treatments as defined by MAP, and 18% as defined by SBP. There was no evidence of patient-specific or universal RBV thresholds that were associated with hypotension. Analysis using the kappa statistic showed that concordance of RBV and hypotension (that is, RBV falling prior to hypotensive episodes rather than rising or remaining stable) was no greater than chance. RBV and rate of change in RBV slopes did not predict hypotension as a dichotomous outcome, and were not correlated with MAP or SBP whether autocorrelation was corrected for or neglected. Substituting rate of change in RBV (RBV slopes) for absolute RBV values did not improve predictive power.
Conclusions: These results suggest that strategies based on online RBV monitoring are unlikely to reduce hypotension in intermittent hemodialysis for ARF that is delivered through central venous catheters.