Rethinking Blood Pressure Targets in Vasodilatory Shock

Recent high-quality evidence shows that targeting higher blood pressures in all patients with vasodilatory shock is associated with increased mortality. This review highlights the need for truly personalised resuscitation strategies. Integrating an individual’s actual pre-morbid blood pressure with bedside dynamic perfusion assessments is likely to result in a reduction of harm caused by excessive vasopressor use and fluid overload.

For decades, the management of vasodilatory shock, especially septic shock, has been anchored to achieving a mean arterial pressure (MAP) threshold, typically more than 65mmHg. While this target remains a current guideline recommendation (Evans et al. 2021), real-world practice has been shown to often exceed this threshold (White et al. 2025). In addition, clinicians perceive lower MAP targets as less acceptable for patients with pre-existing hypertension and acute kidney injury (Young et al. 2025).  This clinical reasoning, however, does not fully account for the core pathophysiology of septic shock, which is characterised by impaired autoregulation, heterogeneous microcirculatory flow, and profound metabolic dysfunction.

Trials investigating higher MAP targets (>75-80mmHg) have shown no mortality benefit and likely harm, potentially via catecholamine-induced toxicity. Evidence from recent randomised controlled trials (Endo et al. 2025) and an updated meta-analysis (Mendes et al. 2025) demonstrates that higher pressure targets can significantly increase mortality without benefiting any subgroup, including the elderly and those with chronic hypertension. This narrative review therefore aims to reconcile bedside habits with contemporary evidence. It summarises the data demonstrating the harm of universally pursuing higher MAP targets, suggests the pathophysiological rationale for this harm, and consolidates the emerging evidence base for a personalised framework.

Rethinking the Goal of Resuscitation

The fundamental objective of shock resuscitation is to restore adequate blood flow and oxygen delivery to tissues, not merely to normalise an arterial pressure value. Yet, in daily practice, blood pressure remains the most visible and easily measurable variable, often shaping bedside decision-making beyond its important albeit limited physiological significance. This clinical practice is not new; since the 1930s blood pressure became an important target because it was readily measurable, whereas blood flow was not (Dünser et al. 2013).

Perfusion depends on the gradient between arterial and venous pressures, tissue metabolic demand and the rheological properties of blood. However, arterial pressure and systemic blood flow are often poorly correlated during critical illness (Wo et al. 1993), with similarly inconsistent coupling at the microcirculatory level (De Backer et al. 2002; Lima et al. 2011). Renal perfusion exemplifies this principle; it depends on the gradient between arterial inflow and venous outflow, not MAP alone, and thereby venous congestion can impair renal blood flow even when MAP appears adequate (Panwar et al. 2025).

The Evolving Evidence Against Universal High MAP Targets

The chronology of evidence against universal high MAP targets reveals a consistent story of a promising hypothesis being systematically disproven. The SEPSISPAM (Asfar et al. 2014) was the first large-scale, randomised controlled trial to evaluate high-MAP directly. It showed no survival advantage for a MAP target of 80-85 mmHg compared to 65-70 mmHg but revealed a higher rate of atrial fibrillation in the high-MAP group. Subsequent studies, such as the 65 trial (Lamontagne et al. 2020), did not observe statistically significant benefits from a permissive hypotensive approach in older patients, though it importantly suggested the safety of a lower target. More recently, the OPTPRESS (Endo et al. 2025) yielded an increased 90-day mortality with higher targets in older patients, with no discernible benefit for those with a history of hypertension, the very subgroup in which benefit was most anticipated.

A recent updated meta-analysis by our group (Mendes et al. 2025) of RCTs in MAP targets in vasodilatory shock showed a 10% relative risk increase in 28-day mortality (95% CI 1.01-1.19; P=0.03) with the universal use of higher MAPs (75-85 mmHg). Bayesian analysis indicated a 98.7% posterior probability of harm for 90-day mortality. Crucially, the higher targets conferred no advantage in any predefined subgroup, including those ≥65 years or with chronic hypertension (Mendes et al. 2025)

When synthesised, these trials reveal a remarkably aligned pattern: a blanket adoption of MAP targets above the 60-70 mmHg threshold does not improve outcomes and may be associated with harm. Mortality appears increased, there is no meaningful improvement in organ dysfunction (e.g., no reduction in renal replacement therapy-free days) and the incidence of atrial fibrillation and other catecholamine-related complications is increased.

The Hidden Cost: Vasopressor Toxicity and Intravenous Fluids

Pursuing high MAP targets incurs a significant cost by means of escalating catecholamine exposure and intravenous fluid administration. Achieving a MAP of 80-85 mmHg often requires from 35 to 110% more noradrenaline (Mendes et al. 2025), which amplifies myocardial oxygen consumption and arrhythmogenic potential (Andreis and Singer 2016).

Beyond these cardiovascular effects, catecholamines may trigger endothelial dysfunction, activate coagulation pathways (Ostrowski et al. 2013) and induce immunosuppression (Stolk et al. 2016), mechanisms that can translate into higher rates of organ failure and mortality (Roberts et al. 2020). Excessive adrenergic stimulation impairs mitochondrial respiration, increases oxidative stress and promotes regional hypoperfusion despite preserved systemic pressure (Lopez Garcia de Lomana et al. 2022). Overall, the idea that pursuing higher blood pressure via exogenous catecholamines improves microcirculatory perfusion-oxygenation is questionable (Boerma and Ince 2010).

Therefore, alongside evidence-based definition of MAP targets, catecholamine-sparing strategies are warranted. A potential strategy is the adjuvant use of vasopressin, which has been shown to reduce  norepinephrine requirements in septic shock (Lajoye et al. 2025). While the evidence of mortality benefit is more nuanced (inconclusive evidence from landmark RCTs including the VANISH (Gordon et al. 2016) and VASST (Russell et al. 2008) trials), computer emulations suggest that early administration of vasopressin may be associated with a modest reduction in mortality, particularly when initiated at lower norepinephrine doses (<0.25 µg/kg/min) and earlier in the course of shock (Kalimouttou et al. 2025). This aligns with findings from Australian (White et al. 2024) and USA (Sacha et al. 2025) multicentre retrospective cohorts that suggest time-dependent benefit from its use. Small physiological studies further suggest that vasopressin may enhance sublingual microcirculatory indices primarily in patients already receiving norepinephrine doses, once again implying that its beneficial effects may stem from catecholamine sparing rather than direct microvascular recruitment (Nascente et al. 2017).

Another risk associated with the pursuit of a higher MAP target is excessive fluid loading. In septic shock, intravenous fluids can paradoxically worsen haemodynamic status and impair perfusion despite transient increases in cardiac output (Byrne et al. 2018). Large boluses raise venous pressures and lead to interstitial oedema, thereby reducing the microvascular driving gradient even when MAP improves (Byrne and Van Haren 2017). Experimental and clinical data indicate that fluid-induced increases in CO often coincide with reduced systemic vascular resistance, underpinning why MAP augmentation is modest and typically short-lived (van Haren et al. 2019).

The haemodynamic consequences of fluid loading and venous hypertension further illustrate the limitations of MAP-centric resuscitation. Because organ perfusion depends on the gradient between arterial inflow and venous outflow, elevations in venous pressure can diminish tissue blood flow even when MAP appears adequate (Panwar et al. 2025). Bedside Doppler assessment with the Venous Excess Ultrasound Score (VExUS) provides a structured approach to quantifying this congestion. In a recent multicentre cohort study, higher VExUS grades were strongly associated with major adverse kidney events at 30 days, highlighting the vulnerability of the kidney to venous hypertension during critical illness (Klompmaker et al. 2025). A contemporary meta-analysis confirmed that VExUS ≥2 significantly increases the risk of acute kidney injury, although no clear effect on all-cause mortality was identified (Melo et al. 2025). These findings show that venous congestion is an independent and clinically meaningful haemodynamic abnormality and highlight the risk of coupling fluid loading with MAP escalation.

Mean Arterial Pressure and the Microcirculation: A Decoupled Relationship

MAP is a composite variable determined by cardiac output and systemic vascular resistance, serving as an indicator of driving pressure rather than a direct measure of capillary flow or cellular oxygen utilisation. In septic shock, this relationship is disrupted by profound dysregulation of vascular tone and blunted autoregulation, which weakens the link between macro-haemodynamic pressure and microcirculatory perfusion (Leone et al. 2015). Consequently, macro-haemodynamic targets like MAP correlate poorly with tissue perfusion, and a pressure-centric approach risks vasopressor overuse without restoring flow (Sanchez et al. 2025). Although microcirculatory parameters offer mechanistic insights, they have not consistently predicted patient outcomes; in a large international study, microcirculatory variables were not independent predictors of mortality, unlike several macrocirculatory parameters (Vellinga et al. 2015).

This fundamental dissociation means that microcirculatory abnormalities often persist despite achieving macro-haemodynamic targets (LeDoux et al. 2000), and the response to interventions targeting MAP is highly variable and dependent on the underlying microvascular pathology (Arnold et al. 2012). For instance, increasing MAP with norepinephrine did not consistently improve sublingual microvascular flow (Dubin et al. 2009), and stepwise elevation from 65 to 85 mmHg produced highly variable effects on organ perfusion (Thooft et al. 2011). This decoupling is further illustrated by the renal response: while one randomised trial found no benefit in urine output or creatinine clearance at 85 mmHg versus 65 mmHg (Bourgoin et al. 2005), a physiological study showed that increasing MAP to 75 mmHg reduced the renal resistive index and improved urine output in most patients, with no further gain at 85 mmHg (Deruddre et al. 2007). The response is also agent-dependent;  norepinephrine yields inconsistent microvascular changes (Jhanji et al. 2009), whereas dobutamine can improve microvascular density and flow independently of its systemic effects (De Backer et al. 2006).

However, the interpretability of these findings is constrained by the studies' limitations: most are small, single-centre physiologic experiments with short assessment periods, providing mechanistic insight but insufficient evidence for long-term, patient-centred outcomes. The variability of their findings underscores the necessity to move from protocols to physiologic individualisation.

From Protocols to Physiology: The Lessons of Individualisation

The evolution of sepsis resuscitation reflects a broader shift from rigid, protocolised care to physiologic reasoning. The Early Goal-Directed Therapy (EGDT) framework emphasised fixed numerical targets for central venous pressure, MAP and central venous oxygen saturation (ScvO₂). However, subsequent large multicentre trials (ProCESS, ARISE, ProMISe) demonstrated that these strict protocols were not superior to clinician-driven care (Munroe et al. 2023). The failure of protocolised EGDT underscored the danger of conflating easily measurable surrogate endpoints with patient-centred outcomes. Furthermore, key vasopressor and aforementioned MAP-target trials vary widely in enrolment timing, illness severity and achieved MAP separation, limiting their applicability to bedside decision-making (Buchtele et al. 2022).

Contemporary approaches now emphasise individualised resuscitation based on dynamic perfusion assessment. The ANDROMEDA-SHOCK (Hernandez et al. 2019) introduced bedside capillary refill time (CRT) as a practical and physiological resuscitation target, showing reduced vasopressor exposure and a strong trend towards lower mortality compared with lactate-guided therapy. Its follow-up trial ANDROMEDA-SHOCK-2 (Andromeda-Shock-2 Investigators for the Andromeda Research Network et al. 2025), which utilised a complex bundle with CRT-guided resuscitation, demonstrated a statistically significant hierarchical composite primary outcome, with a win ratio of 1.16 (95% CI 1.02-1.33; P = .04). Another study testing perfusion-guided resuscitation, the TARTARE-2S (Pettilä et al. 2025), did not achieve statistical significance on its primary or secondary outcomes. Another multicentre RCT evaluating microcirculation-guided resuscitation was early terminated following pre-planned interim analysis meeting futility stopping boundaries (Bruno et al. 2023). In contrast, a single-centre RCT comparing peripheral perfusion-targeted resuscitation versus standard care in sepsis did result in a statistically significant improvement in 30-day mortality (Ferhat et al. 2025).

In summary, recent perfusion-targeted trials have shown promising results albeit without a clear mortality benefit, suggesting that no single variable is the universal "holy grail." Peripheral perfusion index (Ferhat et al. 2025), CRT (Andromeda-Shock-2 Investigators for the Andromeda Research Network et al. 2025; Pettilä et al. 2025) and other clinical markers are potential steps in the right direction, but likely imperfect surrogates for true tissue oxygenation. A planned aggregated data analysis of these perfusion-based resuscitation trials is keenly awaited (PROSPERO CRD420251229564).

Moving Beyond a Uniform MAP Threshold

A growing body of evidence suggests that perfusion pressure deficit rather than absolute MAP values can better identify patients at risk of organ injury during vasodilatory shock. This concept of "relative hypotension," defined by the gap between a patient's achieved MAP and their premorbid autoregulatory range, provides a physiologically grounded alternative to a fixed target. In early observational work, mean perfusion pressure deficit during shock resuscitation was strongly associated with impaired renal perfusion and early organ dysfunction (Panwar et al. 2013). Subsequent synthesis of the data showed that relative hypotension drives kidney injury by lowering renal perfusion pressure even when MAP appears acceptable, and that traditional MAP thresholds may fail to account for the wide inter-individual variability in autoregulation (Panwar 2018).

These concepts were strengthened by prospective multicentre validation, where greater exposure to relative hypotension was independently associated with adverse kidney outcomes and increased mortality (Panwar et al. 2020). Importantly, this relationship persisted even when clinicians maintained MAP ≥65 mmHg, underscoring that "appropriate pressure" must be contextualised to a patient's chronic vascular physiology, particularly previous blood pressure values — rather than solely the label of chronic hypertension (which does not distinguish between untreated hypertension on one end of the spectrum, and well-controlled hypertension on the other end).

This paradigm shift aligns with an emerging view that macro-haemodynamic normalisation should never be assumed to guarantee adequate microcirculatory or organ-level perfusion. Gershengorn (2020) argues that individualisation is not optional but necessary: critically ill patients differ in their vasomotor tone, vascular compliance and organ autoregulatory thresholds, such that fixed MAP targets risk under- or over-resuscitation depending on the patient.

Early feasibility work demonstrates that individualised MAP strategies can be safely implemented at the bedside. A multicentre pilot trial comparing individualised versus standard MAP targets showed that clinicians could reliably titrate vasopressors to individualised goals while maintaining protocol adherence (Panwar et al. 2025). The results of its phase-3 multicentre RCT (NCT05850962) are eagerly expected.

Conclusion

Contemporary evidence demonstrates that universal higher MAP targets for patients with vasodilatory shock are likely associated with increased mortality, potentially mediated via catecholamine toxicity and the adverse effects of fluid overload. The future of shock management may need to involve the integration of multi-modal data and replacing a one-size-fits-all MAP target with a more nuanced, patient-centred approach that prioritises individual characteristics. Trials assessing patients' actual premorbid blood pressure, rather than solely relying on the chronic hypertensive label, to determine personalised MAP targets may add further evidence towards individualised, patient-specific goals (Individualized Blood Pressure Targets Versus Standard Care Among Critically Ill Patients With Shock - A Multicentre Randomised Controlled Trial 2023).

Conflict of Interest

None.