ANDROMEDA-SHOCK 1 to ANDROMEDA-SHOCK 2: Capillary Refill Time & Personalised Haemodynamic Management

From ANDROMEDA-SHOCK 1 to ANDROMEDA-SHOCK 2, capillary refill time evolved from a bedside sign to a personalised, physiology-based resuscitation target with direct implications for clinical practice and ICU organisation.

Introduction

The evolution of haemodynamic resuscitation in septic shock reflects the identity of critical care as a field that balances protocolised pathways with individualised physiology. Few concepts illustrate this transition better than capillary refill time (CRT). Once considered a basic and almost archaic sign, CRT has become central in discussions about personalising resuscitation and assessing tissue perfusion in real time.

The ANDROMEDA-SHOCK programme has been the principal catalyst for this shift. ANDROMEDA-SHOCK 1 (Hernández et al. 2019) demonstrated that peripheral perfusion-targeted resuscitation could outperform lactate-based strategies. ANDROMEDA-SHOCK 2 expanded this concept into a structured, personalised haemodynamic framework driven by physiological cues, fluid responsiveness and rapid functional testing. This article retraces the scientific and organisational path taken between the two trials and explores the practical consequences of CRT-based resuscitation for ICU leaders, clinicians and healthcare systems.

From Simplicity to Structure: The Early Role of CRT

CRT has long been described in the literature, yet it remained peripheral in the practice of modern critical care. ANDROMEDA-SHOCK 1 challenged this marginal role by showing that CRT normalisation paralleled improvements in organ dysfunction, required less fluid and was associated with a higher probability of survival compared with lactate-guided care. The trial signalled a philosophical turning point by demonstrating that microvascular perfusion could be evaluated meaningfully at the bedside without advanced technology. In many middle-income and resource-limited settings, this simplicity carried particular resonance, offering an accessible and cost-neutral tool for guiding resuscitation (Hernández et al. 2020).

The Leap to Personalisation in ANDROMEDA-SHOCK 2

Whilst the first trial placed CRT on the resuscitation map, the second redesigned the entire landscape. ANDROMEDA-SHOCK 2 (Hernández et al. 2025a) was built on the recognition that septic shock is not a single haemodynamic entity, but a syndrome composed of overlapping patterns such as hypovolaemia, vasoplegia and ventricular dysfunction. The trial operationalised this heterogeneity through a personalised haemodynamic resuscitation (PHR) algorithm that used CRT as the primary endpoint, supported by pulse pressure and diastolic arterial pressure as immediate physiological indicators, formal fluid responsiveness evaluation, targeted bedside echocardiography and reversible functional tests such as the mean arterial pressure challenge and low dose dobutamine assessment (Figure 1).

This sequential framework aligned escalation decisions with physiological abnormalities rather than routine or intuition. By focusing on CRT as the anchor variable, ANDROMEDA-SHOCK 2 aimed to avoid unnecessary interventions whilst ensuring that persistent abnormalities triggered an appropriate, stepwise diagnostic and therapeutic response.

Key Findings of ANDROMEDA-SHOCK 2

The trial enrolled 1467 patients from 19 countries and demonstrated a higher probability of "wins" in a hierarchical composite outcome incorporating mortality, duration of vital support and hospital stay. Mortality at 28 days did not differ between groups, but the personalised CRT-guided strategy resulted in more organ support-free days, lower fluid volumes, more frequent CRT normalisation and more physiologically appropriate use of vasopressors and inotropes. These findings emphasise that precision, rather than intensity, improves resuscitation. The reduction in vital support days is particularly relevant for ICU management, where bed availability, staffing pressure and resource constraints are constant challenges.

Implications for ICU Management and Organisation

Implementing CRT-guided personalised resuscitation requires coordinated changes in workflow. Hourly CRT assessments, structured decision-making pathways and systematic escalation create a predictable rhythm for clinicians. Although this may initially appear labour-intensive, the majority of interventions occur early in the algorithm, with only a minority of patients progressing to higher tiers. This structure often reduces downstream workload by preventing fluid overload, avoiding unnecessary interventions and shortening the duration of organ support.

Staff training is central to successful implementation. The trial demonstrated that large numbers of clinicians across diverse healthcare systems can reliably learn CRT measurement and apply physiology-driven algorithms. This suggests that personalised haemodynamic strategies do not depend on advanced technology but on consistent training and culture. For ICU leaders, this represents an opportunity to enhance care quality without major capital investment.

Resource efficiency is another key implication. The CRT-based strategy required fewer fluids, shorter durations of organ support and more selective use of inotropes, all without increasing adverse events. These effects have direct consequences for system-level cost, resource utilisation and staffing logistics.

The approach also illustrates an appealing balance between standardisation and flexibility. The algorithm provides a structured escalation framework whilst allowing clinicians to tailor decisions according to individual physiological patterns. This approach aligns with contemporary ICU priorities, where rigid protocols are gradually giving way to adaptable, physiology-centred models of care that preserve clinical reasoning.

Importantly, CRT-guided resuscitation is feasible in a broad range of healthcare environments. The global participation in ANDROMEDA-SHOCK 2 confirms that essential components of the algorithm—CRT assessment, pulse pressure, diastolic pressure, bedside echocardiography and reversible tests—are widely available. The approach therefore supports greater equity in shock management across diverse resource settings (Hernández et al. 2025a).

How CRT Has Reshaped the Philosophy of Resuscitation

The ANDROMEDA programme has driven a conceptual evolution in septic shock management. CRT has reinforced the principle that perfusion must be prioritised over surrogate macrovascular targets; that interventions must be personalised rather than reflexive; and that rapid, reversible testing offers a pragmatic balance between speed and safety. The structured algorithm helps clinicians navigate physiological complexity whilst minimising unnecessary interventions. Most importantly, the universal accessibility of CRT makes it a tool with global relevance and scalability.

Future Directions

The progression from ANDROMEDA-SHOCK 1 to 2 lays the foundation for future work. Integrating CRT-based strategies with artificial intelligence-supported decision systems could enhance pattern recognition and reduce cognitive load. Further studies are needed to evaluate CRT applicability across diverse skin types and to improve measurement equity. Extending CRT-guided approaches to perioperative shock, trauma or cardiogenic shock may broaden their impact. Understanding the ideal balance between algorithmic structure and clinical judgement remains a crucial area for future refinement. Nursing-led implementation strategies also merit exploration, given the central role of nurses in bedside assessment.

Conclusion

The journey from ANDROMEDA-SHOCK 1 to ANDROMEDA-SHOCK 2 reflects a significant shift in septic shock resuscitation. CRT has evolved from a simple clinical sign to a guiding metric capable of reorganising decision-making, improving efficiency and reducing the burden of organ support. Its accessibility and physiological grounding make it a powerful tool for both high-resource and resource-limited ICUs. As healthcare systems increasingly seek strategies that combine effectiveness, simplicity and cost efficiency, CRT-guided personalised resuscitation stands out as a compelling model for future critical care practice.

Acknowledgment

We extend our gratitude to the entire team involved in ANDROMEDA-SHOCK 1 and 2 for their invaluable support. ORPN and GHP would like to especially acknowledge Karla Ramos, Marian Phinder-Puente, and Gabriela Bautista-Aguilar for their remarkable contributions.

Conflict of Interest

None.