Definition of Refractory Septic Shock: Consensus from SCCM and ESICM

• In Sepsis
• Wed, 25 Mar 2026

A consensus process conducted by the Society of Critical Care Medicine (SCCM) and the European Society of Intensive Care Medicine (ESICM) to establish clinical criteria for defining refractory septic shock, a severe and poorly standardised subset of septic shock, was presented at the SCCM Congress this week.

Sepsis is defined as life-threatening organ dysfunction caused by a dysregulated host response to infection, and refractory septic shock represents its most advanced and severe stage. Despite widespread use of the term, the literature shows considerable variability in how refractory septic shock is defined, particularly regarding hypotension thresholds, lactate levels, and vasopressor dosing. This lack of consistency hinders clinical management, research comparability, and guideline development. The authors aimed to create a unified, consensus-based definition to improve diagnosis, prognostication, research standardisation, and therapeutic decision-making.

To achieve this, a structured Delphi methodology was employed. A multidisciplinary international panel of 56 experts participated in five iterative rounds between May and September 2025. The steering committee initially proposed 34 statements derived from literature review and expert discussion. Consensus was predefined as ≥75% agreement. Over successive rounds, 29 statements were retained and 13 ultimately achieved consensus and stability.

The panel strongly agreed on the need for a comprehensive consensus definition, emphasising its importance for standardised care, improved communication among clinicians, consistent research outcomes, and accurate guideline development. They identified four key domains that should underpin the definition: tissue perfusion, organ dysfunction, fluid resuscitation and intravascular volume status, and use of vasoactive drugs.

Among these, markers of tissue perfusion were central. Serum lactate concentration and capillary refill time (CRT) were selected as key indicators. Persistently elevated lactate reflects impaired tissue perfusion, although it may also result from other causes such as liver dysfunction or metabolic disturbances. CRT, a simple bedside test, was included as an additional surrogate marker, supported by evidence showing improved outcomes when used to guide resuscitation. Notably, no specific lactate cut-off achieved consensus, reflecting ongoing clinical uncertainty.
Fluid management was another essential component. The panel agreed that refractory septic shock should only be diagnosed after adequate initial fluid resuscitation and assessment of fluid responsiveness. Additional fluid should be guided by dynamic reassessment, recognising that haemodynamic responses evolve over time. A patient is considered refractory only when further fluid administration is unlikely to provide benefit.

Vasoactive therapy was a critical defining element. The panel recommended using norepinephrine equivalent dose (NEE) to standardise measurement across different vasopressors. A threshold of >0.5 µg/kg/min NEE was agreed upon as indicative of refractory shock. This reflects severe circulatory failure and is associated with increased mortality risk. Importantly, the total vasopressor dose, rather than the number of agents used, was considered more meaningful.

In terms of diagnostic tools, critical care ultrasonography (CCUS) was the only modality to achieve consensus. It should be used to exclude alternative or mixed causes of shock, such as cardiogenic or obstructive shock, before confirming refractory septic shock. Other diagnostic approaches, including invasive monitoring devices, imaging, and laboratory biomarkers such as troponin or BNP, were not included due to insufficient evidence or lack of specificity.

Several commonly used parameters were explicitly excluded. Central venous oxygen saturation (ScvO₂) was not included due to inconsistent evidence and its dependence on multiple physiological variables. Similarly, urine output was excluded because it is influenced by numerous non-perfusion factors and may not respond immediately to resuscitation. Blood pressure thresholds were also omitted, as patients already meet septic shock criteria and vasopressors often normalise blood pressure despite ongoing perfusion deficits.

The duration of shock and treatment required before labelling it “refractory” was recognised as important but did not reach consensus. Although some studies suggest a window of 6–8 hours, the panel could not agree on a specific timeframe, reflecting variability in clinical practice and patient response.

Based on these findings, the authors propose that refractory septic shock be defined by the presence of persistently elevated lactate and/or prolonged CRT in a patient with septic shock who remains fluid unresponsive, requires norepinephrine equivalent doses greater than 0.5 µg/kg/min, and has undergone CCUS assessment to exclude other causes of shock.

Practical implications of this consensus were also highlighted. Standardised criteria could improve patient stratification in clinical trials, facilitate data aggregation, and enable more targeted therapies for this high-risk population. However, the authors acknowledge that the evidence base remains limited and that these criteria require external validation before widespread clinical implementation.

This establishes a consensus-derived framework for defining refractory septic shock, centred on persistent hypoperfusion, fluid unresponsiveness, high vasopressor requirements, and exclusion of alternative shock states using CCUS. While not yet definitive, these criteria represent an important step towards standardising the identification and management of the most severe form of septic shock.

Source: Critical Care Medicine
Image Credit: iStock