Letters Section Editor: Stephen J. Lurie, MD, PhD, Senior Editor.
To the Editor: Dr Mehta and colleagues1 failed to adjust for many of the significant differences
between patients who did and did not receive diuretics. While the authors
use a propensity score designed to adjust for some differences, they did not
include acute cardiac failure and physiological indicators of advanced cardiac
dysfunction (ie, lower cardiac output and cardiac index, higher pulmonary
artery wedge pressure, and higher systemic vascular resistance in the diuretic
group) into this model or the multivariate logistic regression; they include
only the heart rate (not different between the groups) in the latter. The
inclusion of the history of heart failure in the statistical models does not
adjust for other or more acute cardiac factors that may have had a significant
effect on both diuretic use and mortality.
Štefanec T. Diuretics in Critically Ill Patients With Acute Renal FailureDiuretics in Critically Ill Patients With Acute Renal Failure. JAMA. 2003;289(11):1379-1381. doi:10.1001/jama.289.11.1379-a