Effects of very early start of norepinephrine in patients with septic shock: a propensity score-based analysis

Ospina-Tascón, Gustavo A.; Hernandez, Glenn; Alvarez, Ingrid; Calderón-Tapia, Luis E.; Manzano-Nunez, Ramiro; Sánchez-Ortiz, Alvaro I.; Quiñones, Egardo; Ruiz-Yucuma, Juan E.; Aldana, José L.; Teboul, Jean-Louis; Cavalcanti, Alexandre Biasi; De Backer, Daniel; Bakker, Jan

BACKGROUND: Optimal timing for the start of vasopressors (VP) in septic shock has not been widely studied since it is assumed that fluids must be administered in advance. We sought to evaluate whether a very early start of VP, even without completing the initial fluid loading, might impact clinical outcomes in septic shock. METHODS: A total of 337 patients with sepsis requiring VP support for at least 6 h were initially selected from a prospectively collected database in a 90-bed mixed-ICU during a 24-month period. They were classified into very-early (VE-VPs) or delayed vasopressor start (D-VPs) categories according to whether norepinephrine was initiated or not within/before the next hour of the first resuscitative fluid load. Then, VE-VPs (n = 93) patients were 1:1 propensity matched to D-VPs (n = 93) based on age; source of admission (emergency room, general wards, intensive care unit); chronic and acute comorbidities; and lactate, heart rate, systolic, and diastolic pressure at vasopressor start. A risk-adjusted Cox proportional hazard model was fitted to assess the association between VE-VPs and day 28 mortality. Finally, a sensitivity analysis was performed also including those patients requiring VP support for less than 6 h. RESULTS: Patients subjected to VE-VPs received significantly less resuscitation fluids at vasopressor starting (0[0–510] vs. 1500[650–2300] mL, p < 0.001) and during the first 8 h of resuscitation (1100[500–1900] vs. 2600[1600–3800] mL, p < 0.001), with no significant increase in acute renal failure and/or renal replacement therapy requirements. VE-VPs was related with significant lower net fluid balances 8 and 24 h after VPs. VE-VPs was also associated with a significant reduction in the risk of death compared to D-VPs (HR 0.31, CI95% 0.17–0.57, p < 0.001) at day 28. Such association was maintained after including patients receiving vasopressors for < 6 h. CONCLUSION: A very early start of vasopressor support seems to be safe, might limit the amount of fluids to resuscitate septic shock, and could lead to better clinical outcomes.