β-Blocker Therapy Could Inflate the Diagnosis of Severe Sepsis.

To the Editor: We have read with great interest the article published in a recent issue of Critical Care Medicine, Contenti et al (1) reporting that long-term β-blocker therapy decreases blood lactate concentration in severely septic patients. The authors concluded that excessive plasma catecholamines lead to an acceleration of glycolytic turnover which results in increased lactate production and that this may be precluded by using β-blocker therapy. Although important findings, some issues should be clarified before commenting on the study results. First, we are concerned that the blood pressure lowering effect of β-blocker therapy could inflate the diagnosis of severe sepsis. Indeed, the data of the present study were insufficient to determine whether β-blocker therapy had decreased the blood pressure, and later, severe sepsis was diagnosed or, in fact, the low blood pressure depended on severe illness. Also, previous reports have shown that the halflife of β-blockers ranges from 7 to 16 hours (2, 3). Therefore, a septic patient could be erroneously categorized as having severe sepsis or septic shock as a result of β-blockers’ reversible effect. However, in the present study, the treatment details were unknown. We think that the type, dosage, and time of previous β-blocker therapy are important to interpret the study results. Although the patients previously treated with β-blockers had a high prevalence of heart failure compared with patients who had not taken β-blockers (58.5% vs 27%; p < 0.001), there was no significant difference in vasopressor requirement between the two groups. However, hypotension is a common clinical problem in patients with heart failure (4). Thus, we think that the blood pressure of patients whose heart function was already compromised would be expected to be lower, and these patients would have more need for vasopressor therapy during sepsis. Therefore, a new assessment requires detailed analyses with regard to β-blockers’ reversible effect on blood pressure that weigh whether a proportion of patients with sepsis be categorized as having severe sepsis or septic shock. The authors have disclosed that they do not have any potential conflicts of interest. receiver operating characteristics curve (AUROC) of 0.95 when the change in a flow variable, such as cardiac output or a direct derivative, was measured upon PLR. All included 23 clinical trials indeed used a flow variable as primary parameter. As the maximal effect of PLR occurs within 30–90 seconds in critically ill patients (3, 4), the recommendation to use a fast responding measurement technique enabling a quick prediction of fluid responsiveness in critically ill patients is sensible (5). It is therefore of no surprise that all studies used a fast responding technique, except for the one study by Lakhal et al (1) as rightfully noted by the authors (5).This correction does not affect the observed high diagnostic performance of the fast responding techniques in the meta-analysis and they remain the most frequently used techniques. Nevertheless, it is interesting that the use of a measuring technique requiring several minutes, namely thermodilution, resulted in a comparable AUROC (6). However, in this study, transpulmonary thermodilution required repeated 15 mL fluid injections up to seven boluses, with the authors remarking in their discussion section that “cardiac output measurements using bolus injections during PLR is cumbersome to carry out.” More importantly, since a fluid challenge of only 50–100 mL is capable of predicting fluid responsiveness (7, 8), the obligatory total of thermodilution boluses shed doubt whether the predictive value of PLR, a mini-fluid challenge or a combination was actually observed (6). Furthermore, the fact that fluids need to be administered to predict fluid responsiveness is paradox and undesirable, and reduces the attractive feature of PLR being a “reversible autotransfusion.” All fast responding measurement techniques allow a reliable prediction of fluid responsiveness upon PLR without the necessity of repeated, potentially needless, and harmful fluid administrations. In conclusion, we fully support the recommendation to use fast responding measurement techniques upon PLR to enable a swift and reliable prediction of fluid responsiveness in critically ill patients without the need of potentially deleterious fluid loading. Therefore, we would like to suggest to slightly adjust the title of this letter to “Assessing the effects of passive leg raising: Fast, not Fluids.” The author has disclosed that he does not have any potential conflicts of interest.