LAUSR

Dear Editor, We read with interest the study of Lamontagne et al. [1], but are uncomfortable with their conclusions. Their pooled analysis derives from two randomized, controlled trials (SEPSISPAM [2] and OVATION [3]), and compares 28-day mortality for high and low mean arterial pressure (MAP) targets in patients with shock. Neither of these two studies reported reduced mortality with increased doses of vasopressors, but both suggested greater side effects of this approach in the high MAP target group. In the pooled analysis, the authors report that increasing the MAP target later than 6 h after commencing vasopressor administration increased mortality in all patients. Nevertheless, the SEPSISPAM study showed that increasing the MAP target could prevent kidney failure, but only in the subgroup of patients with a history of hypertension. It seems inappropriate to study the effect of MAP targets on outcomes for all patients, and this analysis would be more pertinent if limited to those with prior hypertension. Moreover, their post hoc analysis of mortality, with inclusions after 6 h on vasopressors, could increase the risk of a type I error: Fig. 1 illustrates the risk of selection bias, while we calculate that ‘only’ four fewer deaths in the high MAP target group would neutralize their result (Fisher test p = 0.07). It is our opinion that, in septic patients with prior hypertension, whether targeting a high MAP more than 6 h after vasopressor initiation is beneficial in terms of kidney protection or detrimental in terms of mortality remains equivocal. We respectfully submit that this hypothesis should be tested before advocating changes to SEPSISPAM’s conclusions.