Influence of Timing and Catecholamine Requirements on Vasopressin Responsiveness in Critically ill Patients with Septic Shock

Introduction: Despite its widespread use, there is a paucity of data to guide the optimal use of arginine vasopressin (AVP) in critically ill patients with septic shock. Methods: This multicenter retrospective cohort study conducted in critically ill adults sought to evaluate the role of catecholamine requirements and timing on responsiveness to AVP. Responsiveness was defined as both a decrease in >= 50% of catecholamine requirements and no decrease in mean arterial pressure (MAP) at 4 hours post-AVP initiation. Primary outcomes of interest included the proportion of patients who started AVP within 4 hours after starting catecholamine therapy, as well as baseline norepinephrine (NE) equivalents (< 15, 15-39, or >= 40 mcg/min). Multivariate analyses and logistic regression were performed to identify other factors associated with AVP responsiveness. Results: There were 300 patients included in this study, with 74 patients being responders and 226 being non-responders. There was no significant difference in the number of patients who received AVP within 4 hours from catecholamine initiation between responders and non-responders (35% vs. 42%, P = 0.29). There were more patients in the non-responder group requiring >= 40 mcg/min of NE equivalents at AVP initiation (30% vs. 16%, P = 0.023). Stress dose steroid use was less common in responders (35% vs. 52%, P = 0.011), which was consistent with logistic regression analysis (OR 0.56, 95% 0.32-0.98, P = 0.044). Clinical outcomes between responders and non-responders were similar, apart from ICU (5.4% vs. 19.5%) and hospital (5.4% vs. 20.4%) mortality being lower in responders (P = 0.0032 and P = 0.0002, respectively). Conclusion: Shorter times to AVP initiation was not associated with responsiveness at 4 hours post-catecholamine initiation, although non-responders tended to require higher doses of NE equivalents at time of AVP initiation. Concomitant corticosteroids were associated with a lower likelihood of AVP responsiveness.