Arginine vasopressin in vasodilatory shock.

To the Editor: Dünser et al1 concluded that the combined infusion of vasopressin and norepinephrine is superior to infusion of norepinephrine alone in the treatment of shock resistant to catecholamine. We believe that this study is flawed and do not support such a conclusion. First of all, the design included two interventions in parallel, one of arginine vasopressin (and its control group in the other arm, without such treatment) and the other with two different doses of norepinephrine. They concluded that the combination group required a lower dose of norepinephrine, which was a consequence of the experimental protocol. The major shortcoming of this study, however, is an evident mistake in the statistical interpretation of their main results. The authors concluded that patients treated with the combination had improved cardiac performance, an effect that did not occur at all. Table 2 of the article shows that the cardiac index (CI) was different between groups among all experiments, a finding that resulted in the P value of 0.001 presented in the table (significant group effect). In fact, the CI was different between groups at the time of the beginning of the infusion, and it remained different for 48 hours. In the ANOVA for multiple measurements and factors, the treatment effect is tested by the time-group interaction, with its corresponding F and P values and not by the group or time effect in isolation. It is clear that there is no time-group interaction in the case of CI and probably not also in the evaluation of mean arterial pressure. On the basis of such findings, the authors stated that norepinephrine is cardiotoxic. Their conclusion is not supported by the reference cited,2 which does not address the potential deleterious effect of norepinephrine in vasodilatory shock. It addresses instead the role of the sympathetic nervous system in the genesis of ventricular arrhythmia.