Standard sepsis algorithms suggest completing a fluid bolus prior to starting vasopressors in patients with septic shock. However, multiple observational trials have shown an association between early vasopressor use and improved outcomes. (Morimatsu 2004; Bai 2014) CENSER is the first RCT to examine the issue…
Permpikul C, Tongyoo S, Viarasilpa T, Trainarongsakul T, Chakorn T, Udompanturak S. Early Use of Norepinephrine in Septic Shock Resuscitation (CENSER) : A Randomized Trial. Am J Respir Crit Care Med. 2019;
CENSER is a single center, double-blind, randomized controlled trial.
Adult patients (18 and older) with a mean arterial blood pressure lower than 65 with infection as a suspected cause.
Exclusions: Patients in septic shock for more than 1 hour, CVA, ACS, acute pulmonary edema, status asthmaticus, active arrhythmias, active GI hemorrhage, pregnancy, seizure, drug overdose, trauma, need for immediate surgery, or advanced staged cancer.
Norepinephrine (0.05 mcg/kg/min or about 3.5 mcg/min for a 70kg adult) run without titration for 24 hours.
Matching placebo (D5W).
Other sepsis management
Both groups received protocol driven sepsis management according to the 2012 surviving sepsis campaign guidelines, including antibiotics, crystalloid bolus, and source control. Fluid therapy was at the discretion of the treating physician. Open label vasopressors were added in both groups if a MAP >65 was not achieved after a 30mL/kg fluid bolus.
The primary outcome was shock control by 6 hours, defined as a MAP >65 on 2 consecutive blood pressures 15 minutes apart plus adequate tissue perfusion, as evidenced by a urine output of 0.5 mL/kg/hr or a 10% decrease in lactate.
They screened 456 patients, and ultimately included 310 in the analysis.
These were sick patients, with an APACHE-II score of 20, mean arterial blood pressure of 56, and a lactate of 2.8. The groups looked very similar at the outset of the trial.
For the primary outcome, shock control by 6 hours, the norepinephrine group looked much better (76.1% vs. 48.4%; odds ratio: 3.4; 95% CI: 2.09–5.53; P<0.001).
The target MAP was achieved in 3.5 hours as compared to 4.75 hours with placebo. Shock control was achieved at 4.75 hours as compared to 6 hours with placebo.
Mortality was 15.5% with norepinephrine versus 21.9% with placebo (relative risk [RR]: 0.79, 95%CI: 0.53-1.11; P=0.15).
There was less cardiogenic edema (14.4% vs 27.7%, p=0.004) and new onset arrhythmias (11% vs 20%, p=0.03) in the norepinephrine group. All other adverse events, including gut and limb ischemia, were similar between the groups. Renal replacement therapy was required in 12% of the norepinephrine group and 14% of the placebo group (p=0.51).
CENSER is an important trial. Current guidelines continue to emphasize relatively large volumes of crystalloid resuscitation before starting vasopressors, but fluids can cause harm. (Levy 2018; Marik 2017) There are 2 observational trials that demonstrate an association between early vasopressor use and improved outcomes, including improved mortality, but observational trials are inherently limited but their various intrinsic biases. (Morimatsu 2004; Bai 2014)
This trial had a number of strengths, including appropriate blinding and allocation concealment, the use of an intention to treat analysis, and complete follow up of their patients. On the other hand, it is a single center trial, and there are hints that practices are somewhat different in Thailand, potentially limiting external validity. Furthermore, the exclusions were broad, and might eliminate a large number of the patients I see with septic shock. Unblinding is likely, as it is hard to mask the jump in BP when starting norepinephrine. Also, they use lactate clearance as a marker of shock resolution, but there is recent evidence that this is not a great marker to guide sepsis resuscitation. (Hernández 2019) (See this EMCrit episode.)
The biggest problem is that the primary outcome is not a patient oriented outcome. Raising blood pressure is a “monitor oriented outcome” or MOO that doesn’t necessarily equate to patient important outcomes. (Raising the blood pressure doesn’t matter much if it comes at the expense of tissue perfusion). They add markers of tissue perfusion, but it isn’t clear how important lactate clearance is. I can buy urine output as a reasonable surrogate, but at the end of the day, it doesn’t matter at all to patients how much they are peeing in the ICU if it doesn’t lead to better outcomes after they leave the ICU.
In this case, patient oriented outcomes might have been better. The mortality number certainly looks better, although the trial wasn’t powered for mortality, and the result isn’t statistically significant. It is reassuring that there weren’t any of the serious side effects we worry about with vasopressors, but the trial is too small to be sure about potential harms.
One interesting note is that norepinephrine was run peripherally if a central line wasn’t placed (which was the case in almost 60% of the cohort). There was 1 case of skin necrosis in both groups, which is clearly not different. Thus, we effectively have a small RCT of norepinephrine used through a peripheral IV to bolster the data we already had indicating that this is a relatively safe practice.
It is also interesting to note the destination of these patients. Despite being in septic shock, and every patient potentially being on a vasopressor, only half of these patients were treated in the ICU. The rest were sent to the ward. Criteria for ICU admission were endotracheal intubation, need for renal replacement therapy, or need for invasive hemodynamic monitoring. We could be overusing ICU resources in Western settings, but this is an important difference that could limit external validity from this single center study in Thailand.
It isn’t time to change sepsis protocols yet, but this is a very promising trial. Early norepinephrine improved some disease oriented outcomes in this RCT, and there were hints that this might result in patient important benefit. I will probably change my practice somewhat, in that I will reach for norepinephrine a little earlier than I have been, but I will wait for the necessary large, multicenter RCT before suggesting this routinely.
Other CENSER FOAMed
Bai X, Yu W, Ji W, Lin Z, Tan S, Duan K, Dong Y, Xu L, Li N. Early versus delayed administration of norepinephrine in patients with septic shock. Critical Care. 2014;18:R532.
Levy MM, Evans LE, Rhodes A. The surviving sepsis campaign bundle: 2018 update. Intensive Care Medicine. Electronic publication ahead of print, PMID 29675566.
Marik PE, Malbrain MLNG. The SEP-1 quality mandate may be harmful: How to drown a patient with 30 ml per kg fluid! Anesthesiology and Intensive Therapy 2017; 49(5) 323-328.
Morimatsu H, Singh K, Uchino S, Bellomo R, Hart G. Early and exclusive use of norepinephrine in septic shock. Resuscitation. 2004;62(2):249–254.
Hernández G, Ospina-Tascón G, Damiani L, et al. Effect of a Resuscitation Strategy Targeting Peripheral Perfusion Status vs Serum Lactate Levels on 28-Day Mortality Among Patients With Septic Shock: The ANDROMEDA-SHOCK Randomized Clinical Trial. JAMA. 2019;321(7):654-664. [PubMed]
Morgenstern, J. CENSER: Early norepinephrine in septic shock, First10EM, March 18, 2019. Available at: