When Does Septic Shock Become Refractory?

Critical care clinicians frequently describe a patient as having refractory septic shock, yet there has never been a universally accepted definition of what that term means. For some clinicians, refractory shock implies escalating vasopressor requirements. Other times, it reflects persistent hyperlactatemia, progressive organ dysfunction, or shock that fails to respond to standard resuscitative measures. Not surprisingly, the literature has been equally inconsistent, with studies employing widely varying thresholds for vasopressor dose, blood pressure targets, lactate levels, duration of shock, and organ dysfunction.¹

This heterogeneity and lack of standardization create a significant problem. If clinicians, investigators, and guideline developers are all using different definitions of refractory septic shock, it becomes difficult to compare studies, identify high-risk patients, develop targeted therapies, or even communicate effectively about the severity of illness. To address this gap, the Society of Critical Care Medicine (SCCM) and the European Society of Intensive Care Medicine (ESICM) recently developed a joint Delphi consensus statement proposing the clinical criteria for refractory septic shock.² Rather than creating a fully validated diagnostic definition, the authors who reviewed the Delphi process sought to establish a practical framework that could standardize communication, guide future research, and serve as a foundation for subsequent validation studies.

The importance of the consensus article extends beyond semantics.² Refractory septic shock represents the subset of patients who are severely ill with septic shock and carries the highest mortality risk. Clinicians often consider rescue therapies or advanced hemodynamic interventions for these patients. Until now, studies investigating therapies for this specific group of patients have often enrolled markedly heterogeneous populations because investigators lacked a common definition. However, a standardized definition may also allow future trials of adjunctive therapies to enroll a more homogeneous population, potentially improving our ability to identify interventions that benefit the most critically ill patients.

The question this consensus attempts to answer is straightforward: at what point should clinicians stop thinking of a patient as having “ordinary” septic shock and begin recognizing a distinct refractory shock phenotype?

Authors of the consensus study assembled a multinational panel of 56 experts representing 22 countries and multiple disciplines, including intensivists, pharmacists, nurses, and researchers.² Through a structured five-round Delphi process, panelists evaluated 34 statements across eight domains relevant to septic shock, including tissue perfusion, fluid resuscitation, vasoactive therapy, blood pressure targets, source control, and adjunctive therapies. Consensus required at least 75% agreement, while stability was assessed across consecutive rounds. Ultimately, the panel reached consensus on 13 statements that formed the basis of the proposed clinical framework.

The resulting framework is remarkably practical. Rather than relying on a single laboratory value or vasopressor threshold, the consensus defines refractory septic shock through four complementary clinical domains:

1. Persistent Tissue Hypoperfusion

The panel identified persistent hyperlactatemia and prolonged capillary refill time (CRT) as the two markers of tissue perfusion that should be incorporated into the definition.² Notably, the panel could not reach consensus on a specific lactate threshold and were divided between the values of 2 mmol/L and 4 mmol/L. Ultimately, while the panel emphasized persistent lactate elevation was a marker of concern, they decided that no single cutoff could adequately capture the complexity of septic shock physiology. The inclusion of CRT is equally noteworthy. This distinction is important because elevated lactate levels should prompt clinical interpretation rather than automatic classification. Once considered an old-fashioned bedside examination maneuver, CRT has re-emerged as a clinically meaningful marker of perfusion. Its inclusion reflects a growing emphasis on bedside assessment of tissue perfusion rather than reliance on laboratory values alone.

2. Lack of Fluid Responsiveness

One of the most clinically relevant aspects of the consensus is its distinction between fluid administration and fluid responsiveness. Rather than recommending endless fluid administration, the authors recommend assessing whether additional fluid is likely to improve perfusion prior to labeling a patient as having refractory septic shock. This is consistent with our growing movement towards dynamic fluid assessments over arbitrary fluid volumes. Whether through passive leg raising, stroke volume assessment, pulse pressure variation, or other validated techniques, clinicians should establish that additional fluid is unlikely to provide meaningful benefits before concluding that shock has become refractory.

3. Significant Vasopressor Requirement

The panel agreed that vasopressor exposure should be quantified using norepinephrine-equivalent dosing and selected a threshold of norepinephrine-equivalent dose >0.5 mg/kg/min.² Clinicians should recognize that this threshold reflects expert consensus rather than a proven biological inflection point. Whether 0.5 mg/kg/min truly identifies a unique patient phenotype remains an important question for future investigation.
 

4. Evaluation for Mixed Shock Physiology

Perhaps the most underappreciated component of the consensus is the recommendation to use critical care ultrasonography when mixed shock is suspected. Before labeling a patient as having refractory septic shock, clinicians should consider alternative or contributing causes of shock. This recommendation acknowledges how many patients with septic shock do not have purely distributive physiologic shock. Identifying a significant cardiogenic or obstructive component may fundamentally alter management and prevent clinicians from incorrectly attributing persistent shock solely to sepsis.

While the components that made the final framework are important, the elements that failed to achieve consensus may be equally instructive for clinicians. Perhaps the most surprising aspect of the consensus is not what was included, but what was excluded.

The exclusion of a mean arterial pressure (MAP) threshold may be the most controversial aspect of the consensus. While hypotension remains central to septic shock, the panel recognized that vasopressors can normalize blood pressure while microcirculatory dysfunction and tissue hypoperfusion persist. In other words, a patient can have an acceptable MAP but may still be inadequately resuscitated. This reflects a growing shift away from blood pressure as the sole marker of shock severity and toward assessment of tissue perfusion.

Central venous oxygen saturation was also excluded because it reflects multiple interacting physiologic variables, including cardiac output, hemoglobin concentration, arterial oxygen saturation, and oxygen consumption. The panel concluded that its interpretation is often complex and inconsistent.

Despite its frequent use as a marker of end-organ perfusion, urine output was excluded since it may lag behind fluid resuscitation and is influenced by numerous factors independent of tissue perfusion, including chronic kidney disease, diuretic exposure, nephrotoxins, and intrinsic renal injury.

Although the panel agreed that organ dysfunction remains central to the concept of refractory septic shock, no specific organ-based criterion achieved sufficient consensus for inclusion in the final framework. In many ways, this finding serves as a reminder that refractory septic shock should be a diagnosis of persistence rather than premature attribution.

Importantly, the consensus study does not introduce a new treatment bundle or management algorithm. Rather, it provides clinicians with a common language for identifying a particularly high-risk subgroup of septic shock patients. Clinicians may reasonably consider a patient to have entered a refractory shock phenotype when persistent tissue hypoperfusion coexists with a lack of fluid responsiveness, significant vasopressor requirements, and an evaluation for alternative shock physiology when clinically indicated. This framework by joint Delphi consensus encourages clinicians to focus on tissue perfusion rather than blood pressure alone, objectively quantify vasopressor burden, assess fluid responsiveness, and actively search for alternative causes of persistent shock.

For decades, refractory septic shock has existed as a concept without a definition. The SCCM/ESICM consensus does not settle every debate, nor does it provide a treatment algorithm. What it does provide is a common language. By shifting the focus from blood pressure alone to tissue perfusion, fluid responsiveness, vasopressor burden, and mixed shock physiology, the authors offer the first broadly accepted framework for identifying patients who are critically ill with septic shock. Whether this definition ultimately improves outcomes remains unknown, but it represents an important first step toward transforming refractory septic shock from a descriptive label into a clinically meaningful phenotype.

References

1. Singer M, Deutschman CS, Seymour CW, et al. The Third International Consensus Definitions for Sepsis and Septic Shock (Sepsis-3). JAMA. 2016;315(8):801-810.
2. Leone M, Myatra SN, Dugar S, et al. Clinical criteria for the definition of refractory septic shock: a joint Delphi consensus from the Society of Critical Care Medicine and European Society of Intensive Care Medicine. Crit Care Med. 2026;54(5):1073-1091.