Reply to El Bèze et al. and to Tung and Crowley

From the Authors:

The CENSER (Early Use of Norepinephrine in Septic Shock Resuscitation) trial examined whether administering low-dose norepinephrine (NE) at the beginning of resuscitation for sepsis-induced hypotension would increase shock control by 6 hours as compared with standard care (1).

After reviewing the results, Tung and Crowley mentioned that the average dose of NE required to achieve the mean arterial pressure goal of 65 mm Hg was 0.1 μg/kg/min, whereas in the intervention arm, the dose of NE was 0.05 μg/kg/min. We agree with this comment. This observation provides not only a hint for the next study but also a possible explanation for the insignificant reduction in the amount of fluid given in the early-NE group. However, certain concerns include the fact that the administration of a higher initial dose of NE requires more careful attention to the peripheral venous site, and placement of the central venous catheter during intravascular volume depletion at the beginning of resuscitation may also be problematic. In their second comment, they noted that the impact of the intervention was not very strong in patients in the 75th percentile. This should be explained by the patients’ heterogeneity. Those who had a serious infection or underlying cardiovascular disease may not have responded well to resuscitation.

In their first comment, El Bèze and colleagues noted that epinephrine was used in our study in a higher proportion of patients than in other studies. It was assumed that the mean dose of NE used before initiating epinephrine was 0.15 μg/kg/min. Actually, our guidelines recommend the use of epinephrine when the NE dose approaches 0.2 μg/kg/min, or when metabolic acidosis is present. The assumed dose was quite close to the recommended one. With regard to their comment about the use of vasopressors during resuscitation in the control group, we note that open-label NE was allowed to start when the blood pressure goal was not reached after optimal fluid resuscitation. The dose adjustment of NE was recommended to be 0.01–0.02 μg/kg/min every 15 minutes. The dosage plots were thus paralleled, as shown in Figure E3B in the online supplement of Reference 1. As for their comment on the sources of sepsis, in which urinary tract infection (UTI) was more prevalent than pneumonia, we note that the sources of infection vary among studies. In our study, the proportions were 23.9–25.8% pneumonia, 29–30.3% UTI, and abdominal infection 20–21.3% (1). In the study by River and colleagues, the proportions were 39.5–38.5% pneumonia, 27.7–25.6% UTI, and 4.2–3.4% abdominal infection (2). Yealy and colleagues reported proportions of 31.9–34.1% pneumonia, 20.2–22.8% UTI, and 11.2–15.7% abdominal infection (3), and Asfar and colleagues’ study had 51.5–52.1% pneumonia, 11.3% UTI, and 16.8–17.3% abdominal infection (4).